DrosEU Phenotyping Working Group Updates

Phenotyping effort

Collaborative phenotyping effort of European Drosophila melanogaster populations

Genomic analyses (Kapun et al. 2020) showed that there is longitudinal population structure, continent-wide sweeps, candidate genes for local climate adaptation by using 48 pooled population samples from 32 locations.

In 2017, DrosEU consortium has decided to make a collaborative phenotyping effort of European Drosophila melanogaster populations at the Groningen meeting.

The call was made with approximate numbers of populations, lines, etc. and under predefined criteria (finish within a year, reps, blocks, etc) and participating labs have indicated which traits they are willing to phenotype.

Sampling Locations

Nine sampling locations were chosen based on genomic data and cover a wide range of latitude (~20°) and longitude (~40°) across the continent (see table below). From each location, 15 to 20 isofemale lines were established in corresponding labs at the sampling location and the isofemale lines were centrally maintained by Élio Sucena at Instituto Gulbenkian de Ciência (IGC), Lisbon, Portugal. A total of 173 isofemale lines were used in this study.

Table X. Sampling locations, coordinates and collectors, sorted by increasing latitude

Country	Location	Latitude	Longitude	Altitude	Collector
Portugal	Recarei	41.150	-8.410	175	Jorge Vieira
Spain	Gimenells (Lleida)	41.618	0.620	173	Josefa Gonzalez
Denmark	Karensminde	55.945	10.213	15	Mads Schou
Germany	Munich	48.180	11.610	520	Amanda Glaser-Schmitt
Austria	Mauternbach	48.375	15.560	572	Andrea Betancourt
Finland	Akaa	61.100	23.520	88	Maaria Kankare
Ukraine	Uman	48.753	30.206	214	Iryna Kozeretska
Turkey	Yesiloz	40.231	32.260	680	Banu Onder
Russia	Valday	57.979	33.244	217	Elena Pasyukova

## Scale on map varies by more than 10%, scale bar may be inaccurate

Contributers (phenotyping)

Contributed labs are Abbott, Bergland, Billeter, Colinet, Flatt, Fricke, Gibert, Gonzalez, Grath, Hoedjes, Kozeretska, Mensch, Onder, Parsch, Pasyukova, Posnien, Ritchie, Schlötterer, Schmidt, Stamenkovic-Radak, Sucena, Tauber, Vieira, Wegener, Zwaan and detailed list below :

Table X. Contributed labs and traits, sorted by PI

Country	Lab	PI	Trait
Sweden	Lund	Abbott	Pigmentation
USA	Charlottesville	Bergland	Diapause
The Netherlands	Groningen	Billeter	Fecundity
France	Rennes	Colinet	Dry weight
Switzerland	Fribourg	Flatt	Diapause
Switzerland	Fribourg	Flatt	Lifespan
Germany	Muenster	Fricke	Fecundity
France	Lyon	Gibert	Development time
France	Lyon	Gibert	Pigmentation
France	Lyon	Gibert	Viability
Spain	Barcelona	Gonzalez	Cold-shock mortality
Spain	Barcelona	Gonzalez	Starvation resistance
Germany	Munich	Grath	Development time
Germany	Munich	Grath	Viability
The Netherlands	Lausanne	Hoedjes	Development time
The Netherlands	Lausanne	Hoedjes	Dry weight
The Netherlands	Lausanne	Hoedjes	Viability
Ukraine	Kyiv	Kozeretska	Cold-shock mortality
Ukraine	Kyiv	Kozeretska	Thorax length
Argentina	Buenos Aiers	Mensch	Chill-coma recovery time
Turkey	Ankara	Onder	Dry weight
Turkey	Ankara	Onder	Starvation resistance
Turkey	Ankara	Onder	Wing area
Germany	Munich	Parsch	Heat-shock mortality
Germany	Munich	Parsch	Lifespan
Russia	Moscow	Pasyukova	Lifespan
Russia	Moscow	Pasyukova	Starvation resistance
Germany	Göttingen	Posnien	Thorax length
Germany	Göttingen	Posnien	Wing area
UK	St Andrews	Ritchie	Thorax length
UK	St Andrews	Ritchie	Wing area
UK	St Andrews	Ritchie	Wing patterning
Austria	Vienna	Schlötterer	Diapause
USA	Philadelphia	Schmidt	Development time
USA	Philadelphia	Schmidt	Pigmentation
USA	Philadelphia	Schmidt	Thorax length
USA	Philadelphia	Schmidt	Time to pupation
USA	Philadelphia	Schmidt	Viability
Serbia	Belgrade	Stamenkovic-Radak	Development time
Serbia	Belgrade	Stamenkovic-Radak	Viability
Serbia	Belgrade	Stamenkovic-Radak	Wing area
Portugal	Lisbon	Sucena	Fly husbandry
Israel	Haifa	Tauber	Locomotor activity
Portugal	Porto	Vieira	Chill-coma recovery time
Portugal	Porto	Vieira	Cold-shock mortality
Portugal	Porto	Vieira	Heat-shock mortality
Germany	Würzburg	Wegener	Circadian eclosion timing
The Netherlands	Wageningen	Zwaan	Development time
The Netherlands	Wageningen	Zwaan	Viability

Contributers (analyses)

Esra Durmaz (all), Envel Kerdaffrec (all), Katja Hoedjes (Via, DT, DW), Banu Onder (SR, WA), Marija Tanaskovic (WA), Chris Wegener (CET), Eran Tauber (LA), Rudolf Rohr (UniFr), coordination-team

Phenotyping

Overview

Methods

Wolbachia

wolbachia_pop_freqs # Wolbachia frequency by country and population abbreviation 

##    Country Population Wolbachia_Freq
## 1  Finland         AK      100.00000
## 2 Portugal         GI       66.66667
## 3  Denmark         KA       85.00000
## 4  Austria         MA       75.00000
## 5  Germany         MU       94.73684
## 6    Spain         RE       47.05882
## 7  Ukraine         UM       69.23077
## 8   Russia         VA       75.00000
## 9   Turkey         YE       85.00000

Diets

Inversion Frequencies

Traits

Contributors and methods for experiments

Viability

Gibert, Grath, Hoedjes, Schmidt, Stamenkovic-Radak and Zwaan Labs

3-5 day old adults (at least 25 pairs) are allowed to lay eggs en masse. Yeast is provided to stimulate egg laying (for at least 2 hours). Eggs are collected, and 40 are placed in each vial¹. Viability is calculated per vial, as the percentage of individuals that emerged from the 40 eggs.

Please see below number of

Developmental time

Egg-to-pupa developmental time

Schmidt Lab

Egg-to-adult developmental time

Gibert, Grath, Hoedjes, Schmidt, Stamenkovic-Radak and Zwaan Labs

Developmental time is scored as both the egg-to-pupa and egg-to-adult development time. Both were scored twice a day, when the chamber lights are turned on and two hours before they are turned off. In order to measure the egg-to-pupa developmental time, the spot where a pupa is found is marked with a permanent marker to keep track of which pupae have emerged in each day. The egg-to-adult developmental time is estimated by counting all emerged adults from the vial, and by sexing them.

Dry weight

Colinet, Hoedjes and Onder Labs

At day 7 after emergence, flies are killed by snap freezing them in liquid nitrogen, by putting them at -20ºC or by putting them into an ethyl acetate solution and stored at -20ºC. Then they are sexed and placed into 96 wells plates, and placed in an oven set at 60-70 °C, for at least 3 days. At this point flies can be stored at room temperature using a protective cover. If this is the case, the day before measurements are made dry flies are again placed for 24h in the oven (60-70 °C) to ensure material is well dehydrated. Flies are then placed on a small piece of aluminium foil for direct weight measurement on microbalance (accuracy 1µg).

Thorax length

Kozeretska, Posnien, Ritchie and Schmidt Labs

Five to seven days old flies are placed onto a double-sided sticky tape attached to a microscope slide and a picture of the thorax taken using a digital camera connected to a dissecting microscope. The same magnification and resolution is always used to increase reproducibility, and a scale bar inserted on each photo to allow transforming pixels into µm units. Thorax length is defined as the distance from the anterior margin of the thorax to the posterior tip of the scutellum and it is measured using the “Straight Line” in ImageJ/Fiji.

Wing area

Onder, Posnien, Ritchie and Stamenkovic-Radak Labs

Both the left and right wings of five to seven days old flies (10 flies per sex per replicate) are removed and placed into a drop of Entellan®Merck, Hoyer’s Medium, sticked to a double-side sticky tape, or taped directly to the slide. Pictures of the wing preparations are taken using a digital camera connected to a dissecting microscope. The same magnification and resolution is always used to increase reproducibility. A scale bar is placed on each photo to allow transforming pixels into µm units. Manual measurements of wing length and wing area are performed using the “Straight Line” and “Polygon Selection” tools, respectively of ImageJ/Fiji (10.1371/journal.pone.0000007).

Fecundity

Billeter and Fricke Labs

For each isofemale line, 10 males and 10 females are placed together in single-sex groups and allowed to mature for five days. Then, they are placed together (5-7 pairs), and mating interactions observed to ensure successful mating (at least 10 min copulation duration) to ensure that we have five successfully mated females. After a successful mating, males are discarded and females allowed to oviposit alone for 48 hours, moved to another vial, and allowed to oviposit for four days, and again moved to another vial and allowed to oviposit for two days to check that of egg-laying stopped. Vials are incubated until all offspring is born. Individuals are then frozen and the offspring counted.

Lifespan

Flatt, Parsch and Pasyukova Labs

Line level lifespan: Ten flies per sex/line are placed in each vial. The age at death will be scored when changing the food, at least three times a week. Five replicates are used.

Population level lifespan: Flies are kept in 1L demography cages (5 flies per line/sex for each population). The age at death will be scored when changing the food, at least three times a week. Ten replicates were performed.

Cold-shock mortality

Gonzalez, Kozeretska and Vieira Labs

Batches of 15-20 seven days old flies are placed for 18 hours in an empty vial immersed in an ice-water slurry box placed at a 4°C room for 18 hours. Then the vials are removed to a bench in a 25°C room and mortality scored 24 hours later.

Chill-coma recovery time

Vieira and Mensh Labs

Sexed flies are placed in an empty vial immersed in an ice-water slurry box placed at a 4°C room in the morning. Six hours later, flies are removed from the tube to individual wells of 24 well plates while being kept on ice. A timer is started once the plate is moved from the ice to a bench in a 25°C room. Each fly is checked by eye for recovery for a maximum of 60 minutes. Flies that are able to stand on their legs are considered recovered and the CCRT (in seconds) recorded.

Heat-shock mortality

Parsch and Vieira Labs

Batches of 15-20 seven days old flies are placed in empty vials inside a 37ºC incubator and mortality checked for 7 hours every 30 minutes.

Diapause

Bergland, Flatt and Schlotterer Labs

In order to induce diapause, two hour old ‘phenotypic virgins’ (pharate or melanized with meconium visible) female flies are exposed to 12°C and 10:14 light/dark hours for 3 weeks, using an incubator that allows temperature tracking. Vials are changed once per week. After three weeks under diapause conditions, flies are frozen at -80C until dissection. Both ovaries will be examined and classified according to the following simplified ‘classification’: 1) < stage 10: diapause; 2) stage 10-13: intermediate; 3) stage 14: non-diapause.

• Standard diapause conditions: 12°C at 10:14 hrs L:D (under these conditions we might expect around 50% dipause frequency)
• duration of exposure: 3 weeks
• assays will be done in incubators (Percival or Memmert); temperature conditions will be tracked using data loggers / iButtons
• Try to have haphazard standardization; similar conditions coupled with consistent tracking of temperature (multiple iButtons, for example)
• We will use egg lays in bottles to be able to collect more flies for the assay (approx. 1-2 bottles per line)
• Larval density will be controlled (approximately) prior to assays: we will do this by letting a fixed number of (approx. 10-20) females (housed together with 5-10 males) oviposit over night (fixed time window) - how many females and the duration of egg laying might need to be adjusted a bit depending on the fertility of the lines (to be determined when lines are being amplified prior to assays)
• The Bergland and Flatt labs will standardize their fly food according to the Gulbenkian recipe (in terms of amounts / proportions); Manolis from the Schlötterer lab will not standardize the food and use their local diet; importantly, flies for the assays will not be exposed to any (additional) live yeast (i.e., the diet will contain liophilized yeast but no live yeast sprinkled on top).
• From the bottles we will collect upon eclosion ‘phenotypic virgins’ (=pharate adults or melanized with meconium visible), approx. within 2 hours after eclosion
• Flies will be kept in vials for assay (15-20 flies in the same vial to account for mortality); we aim to measure a minimum of 10 females per line (from approx. 10-15 lines x 10 populations, yielding about 1000-1500 flies for dissections in total)
• The food in the assays will be changed once per week
• To keep the workload and effort manageable the assays will be performed in 2-3 blocks (optimally only 2) over time: e.g., from each population we will measure a subset of all lines (e.g., 5-7 lines, i.e. 1/2 or 1/3 of all lines) in each of the 2-3 blocks.
• After 3 weeks under diapause conditions, flies will either be dissected or kept frozen until dissection (dry, at -80C; can be stored for a few months)
• We will adopt a simplified diapause ‘classification’:
• Stage 10 will be defined as the checkpoint
• < Stage 10: diapause
• Stage 10-13: intermediate
• Stage 14: non-diapause
• We will examine both ovaries.

Circadian eclosion timing

Wegener Lab

The eclosion rhythmicity has been measured in outcrossed Munich, Akaa, Gimenells, Uman, Mauternbach, Karensminde, Valday, Yesiloz and Recarei strains at:

1. LD14:10 @ 29°C, flies raised at 29°C

2. DD 29°C after LD14:10 entrainment, flies raised at 29°C

3. LD14:10 @ 18°C, flies raised at 18°C

4. DD 18°C after LD14:10 entrainment, flies raised at 18°C

Locomotor activity

Tauber Lab

The locomotor activity of the flies (males only) was measured using the DAM2 Drosophila monitors (Trikinetics Inc., Waltham, MA). Flies were 1-3 days old. Single flies were placed in glass tubes (10 cm × 0.5 cm) that were filled with 2 cm sugar/agar medium. The monitors were placed in light chambers driven by LED, in an incubator at 24°C, ~30% humidity. The flies were entrained to a light-dark cycle (LD 12:12) for 5d and then allowed to free-run for 10 d in constant darkness (DD). The activity data were processed into 30 min bins, and four different variables were analysed. These included the circadian period, and the phase, which were analyzed using the FFT NLLS algorithm available at the BioDare2 server (https://biodare2.ed.ac.uk/). The other two variables, level of activity and the nocturnal/diurnal ratio were analyzed by a custom-made R script.

Starvation resistance

Gonzalez, Onder and Pasyukova Labs

Batches of 10 sexed flies are transferred to glass vials filled with 5 mL of 2% agar for starvation 3-7 days after eclosion. The age at death will be scored every 8 hours.

Pigmentation

Abbott, Gibert and Schmidt

For each line, 10 females, 13-15 days old, either alive or stored in 95% ethanol, are air dried and placed on their left side, and pictures taken using a dissecting microscope. Images are then analysed in ImageJ 1.46r, using the Area Fraction measurement in the Analyze menu. Area Fraction measures the percentage of pixels in a selected area that have been highlighted in red using the Threshold tool. This gives an estimate of the percentage of dark pigmentation on the three terminal tergites of the abdomen (4, 5 and 6).

Data reformatting

MasterSheets

All data files are reformatted as so called Master Sheets , in which the data structure is the same for all the sub-data sets and all details/edits logged in a README sheet.

Table X. Filling in the data spreadsheet and header description

Headers	Descriptions
Supervisor/PI	Lab identifier
Diet	Either standardized S or non-standardized NS
Batch	From 1 to n
Population	Population identifier in abbreviated form (i.e. AK)
Line	Line identifier (i.e. MU1)
Sex	Either female (F) or male (M)
ReplicateVial	Replicate identifier (from 1 to n)
TraitValue	Trait value in x units

## Warning in instance$preRenderHook(instance): It seems your data is too big
## for client-side DataTables. You may consider server-side processing: https://
## rstudio.github.io/DT/server.html

Data check

For each trait, sex and lab, we run counts for population, line and replicate vial (if applicable).

insert count table

Results I : Linear Models

text BLA BLA BLA

Examples & Directories & Legends

Tables

Here is an example code for tables (egg-to-adult viability). The very same code is used for other traits (either with or without Batch) by filtering the data at supervisor, batch, population and/or line levels (and also for sex, if applicable).

## First defined standard error and coefficient of variation

std_err <- function(x) sd(x)/sqrt(length(x))
coef_var <- function(x) sd(x)/mean(x)
estimate_mode <- function(v) {
   uniqv <- unique(v)
   uniqv[which.max(tabulate(match(v, uniqv)))]
}

# Then compute descriptive statistics (mean, median, minimum, maximum, standard error, coefficient of variation and mode) at the defined levels (line with batch information example below) : 

table_Via_Line_wbatch <- write.csv(d_Via %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(ProportionEggtoAdultSurvival), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Viability/table_Via_Line_wbatch.csv", row.names = T)

table_Via_Line_wbatch <- read.csv("Viability/table_Via_Line_wbatch.csv") #file name is here 

Tables for descriptive statistics at population and line levels can be found in the trait directory, under the file name table_TraitAbbreviation_Level_BatchInfo.csv (i.e. Viability/table_Via_Line_wobatch.csv)

Figures

Data range is calculated with #r min() and #r max() functions for each trait.

y-axis on the figures is scaled by the minimum (#r min_trait) and maximum (#r max_trait) values in the full data set for a given trait.

The very same ggplot theme has been used all across the document, called droseu_theme

droseu_theme <- theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black",),  axis.title.x = element_text(size = 16), axis.text.x = element_text(size = 16),axis.text.y = element_text(size = 16),axis.title.y = element_text(size = 16))

p_TraitAbbreviation_Lab + ylim(c(min_trait, max_triat)) + droseu_theme 

Here is an example code for figures (egg-to-adult viabiliy, Gibert Lab). The very same code is used to generate plots for other contributing labs and traits (either for Batch or Population) by filtering the data at supervisor level (for females and males separately, if applicable).

p_Via_Gibert <- ggplot(data = (subset(d_Via,Supervisor.PI=='Gibert')), #subsets for supervisor
                       aes(x=Population_Lat, y=ProportionEggtoAdultSurvival, fill=Batch)) + #define x- and y-axis
                       geom_boxplot(outlier.shape = NA, notch=FALSE) +  #boxplot
                       labs(title="p_Via_Gibert", x="Population", y = "ProportionEggtoAdultSurvival") + 
                      ylim(c(min_Via, max_Via)) + 
                      droseu_theme #y-axis limits, axis labels and droseu_theme

pdf(file="Viability/p_Via_Gibert.pdf",width=8, height=5)

p_Via_Gibert

dev.off()

Linear Models

Here is an example code for linear models (egg-to-adult development time, males, Gibert Lab). The very same code is used for other contributing labs and traits by filtering the data at supervisor level (for females and males separately, if applicable), adding or removing factors if they cause singularity problems (i.e. if only one replicate vial has been phenotyped per line and/or population and/or batch, the lowest level [ replicate_vial in this example] has been dropped from given model).

DT_A_M_lmer_Gibert <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population) + #Line (random) is nested in Population (fixed)
                             (1|Batch) + #Batch is a random effect 
                             (1|ReplicateVial : Line), #Replicate vial (random) is nested in Line (random)
                           data = d_DT_A_M[d_DT_A_M$Supervisor.PI == "Gibert",]) #Filter by Lab in development time, males data 
capture.output(summary(DT_A_M_lmer_Gibert),file = "DevelopmentTime/DT_A_M_lmer_Gibert_sum.txt") #save output summary 
capture.output(anova(DT_A_M_lmer_Gibert),file = "DevelopmentTime/DT_A_M_lmer_Gibert.txt") #save anova output 
capture.output(emmeans(DT_A_M_lmer_Gibert, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file =  "DevelopmentTime/DT_A_M_lmer_Gibert_tk.txt") #run and save post-hoc test (tukey) output 

anova(DT_A_M_lmer_Gibert) #print anova output

summary(DT_A_M_lmer_Gibert) #print model output summary 

Analyses outputs can be found in the trait directory, under the file name TraitAbbreviation_Sex_Function_LabIdentifier.txt (i.e. DevelopmentTime/DT_A_M_lmer_Gibert.txt)

Please note that “Plots and Linear Models by Lab” are presented in alphabetical order.

Viability

dir.create(file.path("Viability"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Gibert Lab :Cristina Vieira, Laurence Mouton, Natacha Kremer, Sonia Martinez, Patricia Gibert

Grath Lab : Ingo Müller, Sonja Grath

Hoedjes Lab : Hristina Kostic, Katja Hoedjes

Schmidt Lab : Ozan Kiratli, Yonatan Babore, Liam Forsythe, Paul Schmidt

Stamenkovic-Radak Lab : Marija Savic Veselinovic, Marija Tanaskovic, Aleksandra Patenkovic, Mihailo Jelic, Katarina Eric, Pavle Eric, Slobodan Davidovic, Marina Stamenkovic-Radak

Zwaan Lab : Joost van den Heuvel, Bas Zwaan

Data reformatting

ifelse(!dir.exists(file.path("Viability")), dir.create(file.path("Viability")), FALSE)

Reading data in R

d_Via <- read.csv("MasterSheets_Oct21_git/VIA_MasterSheet_Oct21.csv")
str(d_Via)

## 'data.frame':    2367 obs. of  12 variables:
##  $ Supervisor.PI               : chr  "Gibert" "Gibert" "Gibert" "Gibert" ...
##  $ Diet                        : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch                       : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population                  : chr  "AK" "AK" "AK" "AK" ...
##  $ Line                        : chr  "AK1" "AK1" "AK1" "AK10" ...
##  $ ReplicateVialOld            : int  1 2 3 1 2 3 1 2 3 1 ...
##  $ ReplicateVial               : chr  "Gibert_1_AK1_1" "Gibert_1_AK1_2" "Gibert_1_AK1_3" "Gibert_1_AK10_1" ...
##  $ ProportionEggtoAdultSurvival: num  0.68 0.73 0.63 0.85 0.75 0.8 0.85 0.88 0.7 0.68 ...
##  $ Country                     : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude                    : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude                   : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude                    : int  88 88 88 88 88 88 88 88 88 88 ...

Factors need reformatting (i.e. Supervisor.PI should be coded as a factor, not character).

str(d_Via)

## 'data.frame':    2367 obs. of  12 variables:
##  $ Supervisor.PI               : chr  "Gibert" "Gibert" "Gibert" "Gibert" ...
##  $ Diet                        : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch                       : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population                  : chr  "AK" "AK" "AK" "AK" ...
##  $ Line                        : chr  "AK1" "AK1" "AK1" "AK10" ...
##  $ ReplicateVialOld            : int  1 2 3 1 2 3 1 2 3 1 ...
##  $ ReplicateVial               : chr  "Gibert_1_AK1_1" "Gibert_1_AK1_2" "Gibert_1_AK1_3" "Gibert_1_AK10_1" ...
##  $ ProportionEggtoAdultSurvival: num  0.68 0.73 0.63 0.85 0.75 0.8 0.85 0.88 0.7 0.68 ...
##  $ Country                     : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude                    : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude                   : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude                    : int  88 88 88 88 88 88 88 88 88 88 ...

d_Via$Supervisor.PI <- as.factor(d_Via$Supervisor.PI)
d_Via$Diet <- as.factor(d_Via$Diet)
d_Via$Batch <- as.factor(d_Via$Batch)
d_Via$Population_Lat <- factor(d_Via$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_Via$Population_Lon <- factor(d_Via$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_Via$Population_Alt <- factor(d_Via$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_Via$Line <- as.factor(d_Via$Line)
d_Via$ReplicateVial <- as.factor(d_Via$ReplicateVial)
d_Via$ProportionEggtoAdultSurvival <- as.numeric(d_Via$ProportionEggtoAdultSurvival)
d_Via$Country <- as.factor(d_Via$Country)
d_Via$Latitude <- as.numeric(d_Via$Latitude)
d_Via$Longitude <- as.numeric(d_Via$Longitude)
d_Via$Altitude <- as.numeric(d_Via$Altitude)

# Now they should be in the correct format, see below. 

str(d_Via)

## 'data.frame':    2367 obs. of  15 variables:
##  $ Supervisor.PI               : Factor w/ 6 levels "Gibert","Grath",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet                        : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch                       : Factor w/ 4 levels "1","2","3","4": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population                  : chr  "AK" "AK" "AK" "AK" ...
##  $ Line                        : Factor w/ 172 levels "AK1","AK10","AK11",..: 1 1 1 2 2 2 4 4 4 9 ...
##  $ ReplicateVialOld            : int  1 2 3 1 2 3 1 2 3 1 ...
##  $ ReplicateVial               : Factor w/ 2367 levels "Gibert_1_AK1_1",..: 1 2 3 4 5 6 7 8 9 10 ...
##  $ ProportionEggtoAdultSurvival: num  0.68 0.73 0.63 0.85 0.75 0.8 0.85 0.88 0.7 0.68 ...
##  $ Country                     : Factor w/ 9 levels "Austria","Denmark",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ Latitude                    : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude                   : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude                    : num  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat              : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon              : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt              : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

# Voila! 

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

Descriptive statistics at the line level, without batch information :

Descriptive statistics at the population level, with batch information :

Descriptive statistics at the population level, without batch information :

Plots and Linear Models by Lab

Overview

min_Via <- min(d_Via$ProportionEggtoAdultSurvival)
max_Via <- max(d_Via$ProportionEggtoAdultSurvival)

y-axis is scaled by the minimum (0) and maximum (1) values in the full data set.

Gibert Lab

model_means_Via_Gibert <- emmeans(object = Via_lmer_Gibert, specs = ~ Population)
model_means_Via_Gibert_cld <- cld(object = model_means_Via_Gibert, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(Via_lmer_Gibert)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value   Pr(>F)   
## Population 37.336  4.1485     9     2  451.29 0.002213 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_Gibert)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## ProportionEggtoAdultSurvival ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch)
##    Data: d_Via[d_Via$Supervisor.PI == "Gibert", ]
## 
## REML criterion at convergence: -687.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.8895 -0.5220  0.0038  0.5549  2.4861 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev.
##  Line:Population (Intercept) 1.064e-02 0.103146
##  Batch           (Intercept) 6.195e-05 0.007871
##  Residual                    9.192e-03 0.095877
## Number of obs: 532, groups:  Line:Population, 169; Batch, 3
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.75693    0.02658 51.26245   28.48   <2e-16 ***
## PopulationGI  0.60921    0.03063 57.25611   19.89   <2e-16 ***
## PopulationKA  0.72271    0.02674 39.22913   27.03   <2e-16 ***
## PopulationMA  0.64415    0.02663 44.78497   24.19   <2e-16 ***
## PopulationMU  0.72064    0.02668 38.85506   27.01   <2e-16 ***
## PopulationRE  0.64691    0.02978 53.90050   21.73   <2e-16 ***
## PopulationUM  0.71452    0.02805 45.74511   25.47   <2e-16 ***
## PopulationVA  0.63614    0.02668 38.85506   23.84   <2e-16 ***
## PopulationYE  0.51092    0.02668 38.85506   19.15   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.034                                                  
## PopulatinKA 0.039  0.037                                           
## PopulatinMA 0.038  0.035  0.041                                    
## PopulatinMU 0.039  0.037  0.043  0.041                             
## PopulatinRE 0.035  0.035  0.038  0.036  0.039                      
## PopulatinUM 0.037  0.035  0.041  0.039  0.041  0.037               
## PopulatinVA 0.039  0.037  0.043  0.041  0.043  0.039  0.041        
## PopulatinYE 0.039  0.037  0.043  0.041  0.043  0.039  0.041  0.043

Grath Lab

model_means_Via_Grath <- emmeans(object = Via_lmer_Grath, specs = ~ Population)
model_means_Via_Grath_cld <- cld(object = model_means_Via_Grath, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(Via_lmer_Grath)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 30.247  10.082     3 27.279  628.46 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_Grath)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: ProportionEggtoAdultSurvival ~ -1 + Population + (1 | Line:Population)
##    Data: d_Via[d_Via$Supervisor.PI == "Grath", ]
## 
## REML criterion at convergence: -155.6
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -3.14211 -0.56745  0.04161  0.55412  2.60033 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.003394 0.05826 
##  Residual                    0.016043 0.12666 
## Number of obs: 147, groups:  Line:Population, 30
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.69027    0.02612 28.41694   26.42   <2e-16 ***
## PopulationMU  0.61603    0.02569 26.74651   23.98   <2e-16 ***
## PopulationRE  0.63583    0.02569 26.74651   24.75   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltMU
## PopulatinMU 0.000        
## PopulatinRE 0.000  0.000

Hoedjes Lab

model_means_Via_Hoedjes <- emmeans(object = Via_lmer_Hoedjes, specs = ~ Population)
model_means_Via_Hoedjes_cld <- cld(object = model_means_Via_Hoedjes, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(Via_lmer_Hoedjes)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 38.302  4.2558     9   158  595.36 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_Hoedjes)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## ProportionEggtoAdultSurvival ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch)
##    Data: d_Via[d_Via$Supervisor.PI == "Hoedjes", ]
## 
## REML criterion at convergence: -712.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.7848 -0.5126 -0.0026  0.5498  2.7446 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev. 
##  Line:Population (Intercept) 1.219e-02 1.104e-01
##  Batch           (Intercept) 2.196e-12 1.482e-06
##  Residual                    7.148e-03 8.455e-02
## Number of obs: 501, groups:  Line:Population, 167; Batch, 4
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## PopulationAK   0.77450    0.02699 157.99983   28.69   <2e-16 ***
## PopulationGI   0.60267    0.03117 157.99983   19.34   <2e-16 ***
## PopulationKA   0.71817    0.02699 157.99983   26.61   <2e-16 ***
## PopulationMA   0.70850    0.02699 157.99983   26.25   <2e-16 ***
## PopulationMU   0.73950    0.02699 157.99983   27.40   <2e-16 ***
## PopulationRE   0.62333    0.03117 157.99983   20.00   <2e-16 ***
## PopulationUM   0.68490    0.02928 157.99983   23.39   <2e-16 ***
## PopulationVA   0.68083    0.02699 157.99983   25.22   <2e-16 ***
## PopulationYE   0.55967    0.02699 157.99983   20.73   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

Schmidt Lab

model_means_Via_Schmidt <- emmeans(object = Via_lm_Schmidt, specs = ~ Population)
model_means_Via_Schmidt_cld <- cld(object = model_means_Via_Schmidt, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(Via_lm_Schmidt) #lm()` is used only for Schmidt Lab's viability data, as only one vial per line was phenotyped 

## Analysis of Variance Table
## 
## Response: ProportionEggtoAdultSurvival
##             Df Sum Sq Mean Sq F value    Pr(>F)    
## Population   9 62.234  6.9148     140 < 2.2e-16 ***
## Residuals  153  7.557  0.0494                      
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lm_Schmidt)

## 
## Call:
## lm(formula = ProportionEggtoAdultSurvival ~ -1 + Population, 
##     data = d_Via[d_Via$Supervisor.PI == "Schmidt", ])
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -0.48129 -0.19237  0.02616  0.16173  0.51871 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## PopulationAK  0.62516    0.04970  12.580  < 2e-16 ***
## PopulationGI  0.48129    0.05738   8.387 3.06e-14 ***
## PopulationKA  0.76034    0.04970  15.300  < 2e-16 ***
## PopulationMA  0.63263    0.05238  12.077  < 2e-16 ***
## PopulationMU  0.57261    0.04970  11.522  < 2e-16 ***
## PopulationRE  0.62982    0.05738  10.976  < 2e-16 ***
## PopulationUM  0.71066    0.05940  11.965  < 2e-16 ***
## PopulationVA  0.58372    0.04970  11.746  < 2e-16 ***
## PopulationYE  0.53773    0.04970  10.821  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.2222 on 153 degrees of freedom
## Multiple R-squared:  0.8917, Adjusted R-squared:  0.8853 
## F-statistic:   140 on 9 and 153 DF,  p-value: < 2.2e-16

p_Via_pop_Schmidt <- ggplot(data = (subset(d_Via,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=ProportionEggtoAdultSurvival, fill=Population)) +geom_boxplot(outlier.shape = NA, notch=FALSE) + labs(title="p_Via_pop_Schmidt", x="Population", y = "ProportionEggtoAdultSurvival") + ylim(c(min_Via, max_Via))+   geom_text(data = model_means_Via_Schmidt_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_Via_pop_Schmidt

Stamenkovic-Radak Lab

model_means_Via_StamenkovicRadak <- emmeans(object = Via_lmer_StamenkovicRadak, specs = ~ Population)
model_means_Via_StamenkovicRadak_cld <- cld(object = model_means_Via_StamenkovicRadak, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(Via_lmer_StamenkovicRadak)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 9.2366  1.0263     9 10.889  106.32 2.508e-09 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_StamenkovicRadak)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## ProportionEggtoAdultSurvival ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch)
##    Data: d_Via[d_Via$Supervisor.PI == "StamenkovicRadak", ]
## 
## REML criterion at convergence: -589.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.6754 -0.5185  0.0052  0.5759  2.2517 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.013176 0.11479 
##  Batch           (Intercept) 0.001207 0.03475 
##  Residual                    0.009652 0.09825 
## Number of obs: 501, groups:  Line:Population, 167; Batch, 4
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.67693    0.03355 21.98005   20.18 1.13e-15 ***
## PopulationGI  0.60257    0.03968 38.88370   15.19  < 2e-16 ***
## PopulationKA  0.64900    0.03349 21.95976   19.38 2.68e-15 ***
## PopulationMA  0.55659    0.03358 21.97873   16.58 6.59e-14 ***
## PopulationMU  0.70096    0.03351 21.96452   20.92 5.41e-16 ***
## PopulationRE  0.62603    0.03645 29.58016   17.17  < 2e-16 ***
## PopulationUM  0.58917    0.03484 25.27417   16.91 2.70e-15 ***
## PopulationVA  0.56078    0.03355 21.98090   16.71 5.55e-14 ***
## PopulationYE  0.49467    0.03349 21.95976   14.77 6.90e-13 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.230                                                  
## PopulatinKA 0.269  0.227                                           
## PopulatinMA 0.272  0.228  0.268                                    
## PopulatinMU 0.270  0.228  0.269  0.271                             
## PopulatinRE 0.247  0.207  0.247  0.248  0.248                      
## PopulatinUM 0.258  0.216  0.259  0.258  0.259  0.239               
## PopulatinVA 0.271  0.232  0.269  0.270  0.270  0.246  0.257        
## PopulatinYE 0.269  0.227  0.269  0.268  0.269  0.247  0.259  0.269

Zwaan Lab

model_means_Via_Zwaan <- emmeans(object = Via_lmer_Zwaan, specs = ~ Population)
model_means_Via_Zwaan_cld <- cld(object = model_means_Via_Zwaan, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(Via_lmer_Zwaan)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 101.61   11.29     9 151.79  517.76 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_Zwaan)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## ProportionEggtoAdultSurvival ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch)
##    Data: d_Via[d_Via$Supervisor.PI == "Zwaan", ]
## 
## REML criterion at convergence: -328.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.8665 -0.4405  0.1045  0.5674  2.7223 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.01052  0.1025  
##  Batch           (Intercept) 0.00000  0.0000  
##  Residual                    0.02180  0.1477  
## Number of obs: 524, groups:  Line:Population, 169; Batch, 2
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## PopulationAK   0.75329    0.02961 148.12936   25.44   <2e-16 ***
## PopulationGI   0.64361    0.03423 147.09939   18.80   <2e-16 ***
## PopulationKA   0.76026    0.02982 150.93460   25.50   <2e-16 ***
## PopulationMA   0.71368    0.02986 152.29083   23.90   <2e-16 ***
## PopulationMU   0.70735    0.03034 158.56478   23.32   <2e-16 ***
## PopulationRE   0.64594    0.03288 144.24242   19.65   <2e-16 ***
## PopulationUM   0.77891    0.03133 149.71959   24.86   <2e-16 ***
## PopulationVA   0.73253    0.03027 157.97927   24.20   <2e-16 ***
## PopulationYE   0.52726    0.03020 158.64348   17.46   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

p_Via_Zwaan

pdf(file="Viability/p_Via_Zwaan.pdf",width=8, height=5)
p_Via_Zwaan
invisible(dev.off())

p_Via_pop_Zwaan

pdf(file="Viability/p_Via_pop_Zwaan.pdf",width=8, height=5)
p_Via_pop_Zwaan
invisible(dev.off())

Lab correlations

There are 3 populations and 29 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

Via_fitted <- combineFitted(labs = c("Gibert", "Grath", "Hoedjes", "Schmidt", "StamenkovicRadak", "Zwaan"), sex = rep("NA", 6), models = list(Via_lmer_Gibert, Via_lmer_Grath, Via_lmer_Hoedjes, Via_lm_Schmidt, Via_lmer_StamenkovicRadak, Via_lmer_Zwaan))

p_Via_lab_cor <- function() scatterPlotMatrix(Via_fitted, sex = "NA")

p_Via_lab_cor()

pdf(file="Viability/p_Via_lab_cor.pdf", width=8, height=8)
p_Via_lab_cor()
invisible(dev.off())

Developmental time

dir.create(file.path("DevelopmentTime"), showWarnings = FALSE)

Egg-to-pupa developmental time

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Schmidt Lab : Paul Schmidt

Data Reformatting

Reading data in R

d_DT_P <- read.csv("MasterSheets_Oct21_git/DTP_MasterSheet_Oct21.csv")
str(d_DT_P) #Factors need reformatting 

## 'data.frame':    3391 obs. of  13 variables:
##  $ Supervisor.PI   : chr  "Schmidt" "Schmidt" "Schmidt" "Schmidt" ...
##  $ Diet            : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch           : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population      : chr  "VA" "VA" "VA" "VA" ...
##  $ Line            : chr  "VA35" "VA35" "VA35" "VA35" ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : chr  "Schmidt_1_VA35_1" "Schmidt_1_VA35_1" "Schmidt_1_VA35_1" "Schmidt_1_VA35_1" ...
##  $ Individual      : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ DT_EggPupa      : int  120 120 120 120 136 136 136 136 136 136 ...
##  $ Country         : chr  "Russia" "Russia" "Russia" "Russia" ...
##  $ Latitude        : num  58 58 58 58 58 ...
##  $ Longitude       : num  33.2 33.2 33.2 33.2 33.2 ...
##  $ Altitude        : int  217 217 217 217 217 217 217 217 217 217 ...

d_DT_P$Supervisor.PI <- as.factor(d_DT_P$Supervisor.PI)
d_DT_P$Diet <- as.factor(d_DT_P$Diet)
d_DT_P$Batch <- as.factor(d_DT_P$Batch)
d_DT_P$Population_Lat <- factor(d_DT_P$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_DT_P$Population_Lon <- factor(d_DT_P$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_DT_P$Population_Alt <- factor(d_DT_P$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_DT_P$Line <- as.factor(d_DT_P$Line)
d_DT_P$ReplicateVial <- as.factor(d_DT_P$ReplicateVial)
d_DT_P$DT_EggPupa <- as.numeric(d_DT_P$DT_EggPupa)

str(d_DT_P) #They should be ok now. 

## 'data.frame':    3391 obs. of  16 variables:
##  $ Supervisor.PI   : Factor w/ 1 level "Schmidt": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet            : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch           : Factor w/ 1 level "1": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population      : chr  "VA" "VA" "VA" "VA" ...
##  $ Line            : Factor w/ 161 levels "AK1","AK10","AK11",..: 131 131 131 131 131 131 131 131 131 131 ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : Factor w/ 161 levels "Schmidt_1_AK1_1",..: 131 131 131 131 131 131 131 131 131 131 ...
##  $ Individual      : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ DT_EggPupa      : num  120 120 120 120 136 136 136 136 136 136 ...
##  $ Country         : chr  "Russia" "Russia" "Russia" "Russia" ...
##  $ Latitude        : num  58 58 58 58 58 ...
##  $ Longitude       : num  33.2 33.2 33.2 33.2 33.2 ...
##  $ Altitude        : int  217 217 217 217 217 217 217 217 217 217 ...
##  $ Population_Lat  : Factor w/ 9 levels "YE","RE","GI",..: 8 8 8 8 8 8 8 8 8 8 ...
##  $ Population_Lon  : Factor w/ 9 levels "RE","GI","KA",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Alt  : Factor w/ 9 levels "KA","AK","GI",..: 6 6 6 6 6 6 6 6 6 6 ...

Descriptive Statistics

# Note that the trait has been phenotyped only in Schmidt lab and in one batch.

Descriptive statistics at the line level :

Descriptive statistics at the population level:

Plots and Linear Models by Lab

min_DT_P <- min(d_DT_P$DT_EggPupa)
max_DT_P <- max(d_DT_P$DT_EggPupa)

y-axis is scaled by the minimum (96) and maximum (192) values in the full data set.

Schmidt Lab

DT_P_lmer <- lmer(DT_EggPupa ~ -1 + Population + (1|Population:Line), data = d_DT_P)
capture.output(summary(DT_P_lmer),file = "DevelopmentTime/DT_P_lmer_sum.txt")
capture.output(anova(DT_P_lmer),file = "DevelopmentTime/DT_P_F_lmer.txt")
capture.output(emmeans(DT_P_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_P_tk.txt")

model_means_DT_P_Schmidt <- emmeans(object = DT_P_lmer, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 3391' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 3391)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 3391' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 3391)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_P_Schmidt_cld <- cld(object = model_means_DT_P_Schmidt, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_P_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2027827  225314     9 147.82  1892.9 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_P_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggPupa ~ -1 + Population + (1 | Population:Line)
##    Data: d_DT_P
## 
## REML criterion at convergence: 26303.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.9734 -0.6092 -0.0831  0.3912  4.7796 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Population:Line (Intercept) 158.4    12.59   
##  Residual                    119.0    10.91   
## Number of obs: 3391, groups:  Population:Line, 161
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  127.502      2.867 143.578   44.48   <2e-16 ***
## PopulationGI  140.422      3.742 167.058   37.53   <2e-16 ***
## PopulationKA  127.292      2.883 145.919   44.16   <2e-16 ***
## PopulationMA  132.004      2.956 146.299   44.66   <2e-16 ***
## PopulationMU  129.730      2.865 143.161   45.29   <2e-16 ***
## PopulationRE  135.270      3.347 149.396   40.42   <2e-16 ***
## PopulationUM  131.161      3.299 141.662   39.76   <2e-16 ***
## PopulationVA  140.765      2.966 148.082   47.47   <2e-16 ***
## PopulationYE  135.200      2.890 148.107   46.77   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_DT_P_Schmidt <- ggplot(data = (subset(d_DT_P,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=DT_EggPupa, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_P_Schmidt", x="Population", y = "DT_EggPupa") + ylim(c(min_DT_P, max_DT_P))+   geom_text(data = model_means_DT_P_Schmidt_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_DT_P_Schmidt

pdf(file="DevelopmentTime/p_DT_P_Schmidt.pdf",width=8, height=5)
p_DT_P_Schmidt
invisible(dev.off())

Egg-to-adult developmental time

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Gibert Lab : Cristina Vieira, Laurence Mouton, Natacha Kremer, Sonia Martinez, Patricia Gibert

Grath Lab : Ingo Müller, Sonja Grath

Hoedjes Lab : Hristina Kostic, Katja Hoedjes

Schmidt Lab : Ozan Kiratli, Yonatan Babore, Liam Forsythe, Paul Schmidt

Stamenkovic-Radak Lab : Marija Savic Veselinovic, Marija Tanaskovic, Aleksandra Patenkovic, Mihailo Jelic, Katarina Eric, Pavle Eric, Slobodan Davidovic, Marina Stamenkovic-Radak

Zwaan Lab : Joost van den Heuvel, Bas Zwaan

Data Reformatting

d_DT_A <- read.csv("MasterSheets_Oct21_git/DTA_MasterSheet_Oct21.csv")
str(d_DT_A) #Factors need reformatting 

## 'data.frame':    57609 obs. of  14 variables:
##  $ Supervisor.PI   : chr  "Gibert" "Gibert" "Gibert" "Gibert" ...
##  $ Diet            : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch           : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Sex             : chr  "F" "F" "F" "F" ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : chr  "Gibert_1_AK1_1" "Gibert_1_AK1_1" "Gibert_1_AK1_1" "Gibert_1_AK1_1" ...
##  $ Individual      : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ DT_EggAdult     : num  202 202 202 202 202 202 202 202 202 202 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...

d_DT_A$Supervisor.PI <- as.factor(d_DT_A$Supervisor.PI)
d_DT_A$Diet <- as.factor(d_DT_A$Diet)
d_DT_A$Batch <- as.factor(d_DT_A$Batch)
d_DT_A$Population_Lat <- factor(d_DT_A$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_DT_A$Population_Lon <- factor(d_DT_A$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_DT_A$Population_Alt <- factor(d_DT_A$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_DT_A$Line <- as.factor(d_DT_A$Line)
d_DT_A$Sex <- as.factor(d_DT_A$Sex)
d_DT_A$ReplicateVial <- as.factor(d_DT_A$ReplicateVial)
d_DT_A$DT_EggAdult <- as.numeric(d_DT_A$DT_EggAdult)

str(d_DT_A) #They should be ok now. 

## 'data.frame':    57609 obs. of  17 variables:
##  $ Supervisor.PI   : Factor w/ 6 levels "Gibert","Grath",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet            : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch           : Factor w/ 4 levels "1","2","3","4": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : Factor w/ 171 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex             : Factor w/ 2 levels "F","M": 1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : Factor w/ 2300 levels "Gibert_1_AK1_1",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Individual      : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ DT_EggAdult     : num  202 202 202 202 202 202 202 202 202 202 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat  : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon  : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt  : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

#Create data sheets for females (d_DT_A_F) and males (d_DT_A_M), by sorting at sex level.
d_DT_A_F <-subset(d_DT_A,Sex=='F')
d_DT_A_M <-subset(d_DT_A,Sex=='M')

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

Descriptive statistics at the line level, without batch information :

Descriptive statistics at the population level, with batch information :

Descriptive statistics at the population level, without batch information :

Plots and Linear Models by Lab

Overview

min_DT_A <- min(d_DT_A$DT_EggAdult)
max_DT_A <- max(d_DT_A$DT_EggAdult)

y-axis is scaled by the minimum (150) and maximum (394) values in the full data set.

p_DT_A_F_all <- ggplot(data = d_DT_A_F, aes(x=Population_Lat, y=DT_EggAdult, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_all", x="Population", y = "DW_micrograms") + ylim(c(min_DT_A, max_DT_A))+ droseu_theme

## quartz_off_screen 
##                 2

p_DT_A_M_all <- ggplot(data = d_DT_A_M, aes(x=Population_Lat, y=DT_EggAdult, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_all", x="Population", y = "DW_micrograms") + ylim(c(min_DT_A, max_DT_A))+ droseu_theme

## quartz_off_screen 
##                 2

Gibert Lab

DT_A_F_lmer_Gibert <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population) + (1|Batch) + (1|ReplicateVial : Line), data = d_DT_A_F[d_DT_A_F$Supervisor.PI == "Gibert",])
capture.output(summary(DT_A_F_lmer_Gibert),file = "DevelopmentTime/DT_A_F_lmer_Gibert_sum.txt")
capture.output(anova(DT_A_F_lmer_Gibert),file = "DevelopmentTime/DT_A_F_lmer_Gibert.txt")
capture.output(emmeans(DT_A_F_lmer_Gibert, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"), file = "DevelopmentTime/DT_A_F_lmer_Gibert_tk.txt")

model_means_DT_A_F_Gibert <- emmeans(object = DT_A_F_lmer_Gibert, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 6817' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 6817)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 6817' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 6817)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_A_F_Gibert_cld <- cld(object = model_means_DT_A_F_Gibert, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_F_lmer_Gibert)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 5881228  653470     9     2  6961.1 0.0001436 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_F_lmer_Gibert)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: d_DT_A_F[d_DT_A_F$Supervisor.PI == "Gibert", ]
## 
## REML criterion at convergence: 50925
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.4211 -0.5914 -0.1334  0.4381  5.9468 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept)  4.727   2.174   
##  Line:Population    (Intercept) 36.111   6.009   
##  Batch              (Intercept)  1.175   1.084   
##  Residual                       93.874   9.689   
## Number of obs: 6817, groups:  
## ReplicateVial:Line, 531; Line:Population, 169; Batch, 3
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  222.380      1.571  19.336   141.6   <2e-16 ***
## PopulationGI  226.716      1.810  23.726   125.3   <2e-16 ***
## PopulationKA  223.358      1.604  15.418   139.3   <2e-16 ***
## PopulationMA  222.090      1.608  15.561   138.1   <2e-16 ***
## PopulationMU  223.687      1.603  15.387   139.6   <2e-16 ***
## PopulationRE  227.852      1.771  22.073   128.7   <2e-16 ***
## PopulationUM  225.486      1.670  17.868   135.0   <2e-16 ***
## PopulationVA  223.889      1.610  15.644   139.1   <2e-16 ***
## PopulationYE  217.802      1.630  16.455   133.6   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.183                                                  
## PopulatinKA 0.206  0.197                                           
## PopulatinMA 0.205  0.196  0.222                                    
## PopulatinMU 0.206  0.197  0.223  0.223                             
## PopulatinRE 0.187  0.180  0.201  0.200  0.202                      
## PopulatinUM 0.197  0.189  0.214  0.214  0.214  0.193               
## PopulatinVA 0.205  0.196  0.222  0.222  0.222  0.201  0.213        
## PopulatinYE 0.202  0.194  0.219  0.219  0.219  0.198  0.211  0.219

DT_A_M_lmer_Gibert <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population) + (1|Batch) + (1|ReplicateVial : Line), data = d_DT_A_M[d_DT_A_M$Supervisor.PI == "Gibert",])
capture.output(summary(DT_A_M_lmer_Gibert),file = "DevelopmentTime/DT_A_M_lmer_Gibert_sum.txt")
capture.output(anova(DT_A_M_lmer_Gibert),file = "DevelopmentTime/DT_A_M_lmer_Gibert.txt")
capture.output(emmeans(DT_A_M_lmer_Gibert, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_M_lmer_Gibert_tk.txt")

model_means_DT_A_M_Gibert <- emmeans(object = DT_A_M_lmer_Gibert, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 6566' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 6566)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 6566' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 6566)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_A_M_Gibert_cld <- cld(object = model_means_DT_A_M_Gibert, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_M_lmer_Gibert)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 10876506 1208501     9     2   11334 8.823e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_M_lmer_Gibert)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: d_DT_A_M[d_DT_A_M$Supervisor.PI == "Gibert", ]
## 
## REML criterion at convergence: 49809.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.4032 -0.6057 -0.1597  0.4187  5.0802 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept)   3.6928  1.9217 
##  Line:Population    (Intercept)  31.0297  5.5704 
##  Batch              (Intercept)   0.6211  0.7881 
##  Residual                       106.6292 10.3261 
## Number of obs: 6566, groups:  
## ReplicateVial:Line, 533; Line:Population, 169; Batch, 3
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  226.515      1.412  30.257   160.4   <2e-16 ***
## PopulationGI  231.876      1.634  36.517   141.9   <2e-16 ***
## PopulationKA  227.223      1.437  23.884   158.1   <2e-16 ***
## PopulationMA  225.300      1.445  24.338   156.0   <2e-16 ***
## PopulationMU  227.801      1.434  23.728   158.8   <2e-16 ***
## PopulationRE  228.584      1.583  33.036   144.4   <2e-16 ***
## PopulationUM  229.773      1.501  27.748   153.0   <2e-16 ***
## PopulationVA  228.383      1.441  24.170   158.5   <2e-16 ***
## PopulationYE  223.520      1.464  25.700   152.7   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.120                                                  
## PopulatinKA 0.136  0.130                                           
## PopulatinMA 0.135  0.128  0.148                                    
## PopulatinMU 0.136  0.130  0.148  0.148                             
## PopulatinRE 0.124  0.120  0.134  0.133  0.134                      
## PopulatinUM 0.130  0.124  0.142  0.141  0.142  0.128               
## PopulatinVA 0.135  0.129  0.148  0.147  0.148  0.134  0.141        
## PopulatinYE 0.133  0.127  0.145  0.145  0.146  0.132  0.139  0.145

p_DT_A_F_Gibert <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_Gibert", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ droseu_theme

p_DT_A_F_Gibert

pdf(file="DevelopmentTime/p_DT_A_F_Gibert.pdf",width=8, height=5)
p_DT_A_F_Gibert
dev.off()

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

p_DT_A_M_pop_Gibert <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_pop_Gibert", x="Population", y = "DT_EggAdult")+ ylim(c(min_DT_A, max_DT_A))+  geom_text(data = model_means_DT_A_M_Gibert_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_DT_A_M_pop_Gibert

pdf(file="DevelopmentTime/p_DT_A_M_pop_Gibert.pdf",width=8, height=5)
p_DT_A_M_pop_Gibert
dev.off()

## quartz_off_screen 
##                 2

Grath Lab

DT_A_F_lmer_Grath <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population), data = d_DT_A_F[d_DT_A_F$Supervisor.PI == "Grath",])
capture.output(summary(DT_A_F_lmer_Grath),file = "DevelopmentTime/DT_A_F_lmer_Grath_sum.txt")
capture.output(anova(DT_A_F_lmer_Grath),file = "DevelopmentTime/DT_A_F_lmer_Grath.txt")
capture.output(emmeans(DT_A_F_lmer_Grath, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_F_lmer_Grath_tk.txt")

model_means_DT_A_F_Grath <- emmeans(object = DT_A_F_lmer_Grath, specs = ~ Population)
model_means_DT_A_F_Grath_cld <- cld(object = model_means_DT_A_F_Grath, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_F_lmer_Grath)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 22182190 7394063     3 15.388   23537 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_F_lmer_Grath)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population)
##    Data: d_DT_A_F[d_DT_A_F$Supervisor.PI == "Grath", ]
## 
## REML criterion at convergence: 7659.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.0678 -0.7463  0.0894  0.5972  4.2158 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  22.4     4.732  
##  Residual                    314.1    17.724  
## Number of obs: 890, groups:  Line:Population, 23
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  309.626      1.722  13.465  179.80   <2e-16 ***
## PopulationMU  305.631      1.756  14.520  174.06   <2e-16 ***
## PopulationRE  321.616      3.599  19.544   89.37   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltMU
## PopulatinMU 0.000        
## PopulatinRE 0.000  0.000

DT_A_M_lmer_Grath <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population), data = d_DT_A_M[d_DT_A_M$Supervisor.PI == "Grath",])
capture.output(summary(DT_A_M_lmer_Grath),file = "DevelopmentTime/DT_A_M_lmer_Grath_sum.txt")
capture.output(anova(DT_A_M_lmer_Grath),file = "DevelopmentTime/DT_A_M_lmer_Grath.txt")
capture.output(emmeans(DT_A_M_lmer_Grath, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_M_lmer_Grath_tk.txt")

model_means_DT_A_M_Grath <- emmeans(object = DT_A_M_lmer_Grath, specs = ~ Population)
model_means_DT_A_M_Grath_cld <- cld(object = model_means_DT_A_M_Grath, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_M_lmer_Grath)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 12388975 4129658     3 18.297   13725 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_M_lmer_Grath)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population)
##    Data: d_DT_A_M[d_DT_A_M$Supervisor.PI == "Grath", ]
## 
## REML criterion at convergence: 7557.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.4732 -0.5678 -0.0897  0.6034  4.2694 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  47.61    6.90   
##  Residual                    300.89   17.35   
## Number of obs: 881, groups:  Line:Population, 23
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  317.567      2.349  17.582   135.2   <2e-16 ***
## PopulationMU  312.148      2.360  17.944   132.2   <2e-16 ***
## PopulationRE  324.298      4.412  19.486    73.5   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltMU
## PopulatinMU 0.000        
## PopulatinRE 0.000  0.000

pdf(file="DevelopmentTime/p_DT_A_F_Grath.pdf",width=8, height=5)
p_DT_A_F_Grath <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Grath')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_Grath", x="Population", y = "DT_EggAdult")

p_DT_A_F_Grath + ylim(c(min_DT_A, max_DT_A))+ droseu_theme
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_F_pop_Grath <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Grath')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_pop_Grath", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ geom_text(data = model_means_DT_A_F_Grath_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_DT_A_F_pop_Grath

pdf(file="DevelopmentTime/p_DT_A_F_pop_Grath.pdf",width=8, height=5)
p_DT_A_F_pop_Grath
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_M_Grath <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Grath')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_Grath", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ droseu_theme

p_DT_A_M_Grath

pdf(file="DevelopmentTime/p_DT_A_M_Grath.pdf",width=8, height=5)
p_DT_A_M_Grath
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_M_pop_Grath <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Grath')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_pop_Grath", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ geom_text(data = model_means_DT_A_M_Grath_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_DT_A_M_pop_Grath

pdf(file="DevelopmentTime/p_DT_A_M_pop_Grath.pdf",width=8, height=5)
p_DT_A_M_pop_Grath
dev.off()

## quartz_off_screen 
##                 2

Hoedjes Lab

DT_A_F_lmer_Hoedjes <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population) + (1|Batch) + (1|ReplicateVial : Line), data = d_DT_A_F[d_DT_A_F$Supervisor.PI == "Hoedjes",])
capture.output(summary(DT_A_F_lmer_Hoedjes),file = "DevelopmentTime/DT_A_F_lmer_Hoedjes_sum.txt")
capture.output(anova(DT_A_F_lmer_Hoedjes),file = "DevelopmentTime/DT_A_F_lmer_Hoedjes.txt")
capture.output(emmeans(DT_A_F_lmer_Hoedjes, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_F_lmer_Hoedjes_tk.txt")

model_means_DT_A_F_Hoedjes <- emmeans(object = DT_A_F_lmer_Hoedjes, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 6939' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 6939)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 6939' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 6939)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_A_F_Hoedjes_cld <- cld(object = model_means_DT_A_F_Hoedjes, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_F_lmer_Hoedjes)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 16156401 1795156     9 12.393   11812 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_F_lmer_Hoedjes)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: d_DT_A_F[d_DT_A_F$Supervisor.PI == "Hoedjes", ]
## 
## REML criterion at convergence: 54971.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.7059 -0.5552 -0.1930  0.3779 12.7479 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept)   4.685   2.164  
##  Line:Population    (Intercept)  21.642   4.652  
##  Batch              (Intercept)   1.381   1.175  
##  Residual                       151.975  12.328  
## Number of obs: 6939, groups:  
## ReplicateVial:Line, 501; Line:Population, 167; Batch, 4
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  231.278      1.290  30.041   179.3   <2e-16 ***
## PopulationGI  237.944      1.486  47.280   160.1   <2e-16 ***
## PopulationKA  231.098      1.298  30.799   178.1   <2e-16 ***
## PopulationMA  233.211      1.303  31.308   179.0   <2e-16 ***
## PopulationMU  229.501      1.296  30.621   177.1   <2e-16 ***
## PopulationRE  237.728      1.487  45.160   159.9   <2e-16 ***
## PopulationUM  232.195      1.398  38.781   166.2   <2e-16 ***
## PopulationVA  229.880      1.306  31.587   176.0   <2e-16 ***
## PopulationYE  233.177      1.322  33.122   176.4   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.180                                                  
## PopulatinKA 0.206  0.179                                           
## PopulatinMA 0.205  0.178  0.204                                    
## PopulatinMU 0.207  0.179  0.205  0.204                             
## PopulatinRE 0.180  0.165  0.179  0.178  0.179                      
## PopulatinUM 0.191  0.171  0.190  0.190  0.191  0.177               
## PopulatinVA 0.205  0.178  0.204  0.203  0.204  0.178  0.189        
## PopulatinYE 0.202  0.175  0.201  0.200  0.202  0.175  0.187  0.200

DT_A_M_lmer_Hoedjes <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population) + (1|Batch) + (1|ReplicateVial : Line), data = d_DT_A_M[d_DT_A_M$Supervisor.PI == "Hoedjes",])
capture.output(summary(DT_A_M_lmer_Hoedjes),file = "DevelopmentTime/DT_A_M_lmer_Hoedjes_sum.txt")
capture.output(anova(DT_A_M_lmer_Hoedjes),file = "DevelopmentTime/DT_A_M_lmer_Hoedjes.txt")
capture.output(emmeans(DT_A_M_lmer_Hoedjes, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_M_lmer_Hoedjes_tk.txt")

model_means_DT_A_M_Hoedjes <- emmeans(object = DT_A_M_lmer_Hoedjes, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 6649' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 6649)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 6649' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 6649)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_A_M_Hoedjes_cld <- cld(object = model_means_DT_A_M_Hoedjes, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_M_lmer_Hoedjes)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 14164896 1573877     9 11.733  7669.4 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_M_lmer_Hoedjes)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: d_DT_A_M[d_DT_A_M$Supervisor.PI == "Hoedjes", ]
## 
## REML criterion at convergence: 54596.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -6.8700 -0.5946 -0.0822  0.3611 10.9394 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept)   3.107   1.763  
##  Line:Population    (Intercept)  24.219   4.921  
##  Batch              (Intercept)   2.565   1.602  
##  Residual                       205.215  14.325  
## Number of obs: 6649, groups:  
## ReplicateVial:Line, 501; Line:Population, 167; Batch, 4
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  238.685      1.466  19.775   162.8   <2e-16 ***
## PopulationGI  244.647      1.656  30.261   147.8   <2e-16 ***
## PopulationKA  236.153      1.470  19.993   160.6   <2e-16 ***
## PopulationMA  239.367      1.465  19.701   163.4   <2e-16 ***
## PopulationMU  236.531      1.466  19.789   161.3   <2e-16 ***
## PopulationRE  244.438      1.667  30.219   146.6   <2e-16 ***
## PopulationUM  237.762      1.568  24.969   151.6   <2e-16 ***
## PopulationVA  235.826      1.473  20.143   160.1   <2e-16 ***
## PopulationYE  239.204      1.502  21.765   159.2   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.264                                                  
## PopulatinKA 0.298  0.263                                           
## PopulatinMA 0.299  0.264  0.298                                    
## PopulatinMU 0.298  0.264  0.297  0.299                             
## PopulatinRE 0.262  0.241  0.261  0.263  0.262                      
## PopulatinUM 0.279  0.252  0.278  0.279  0.279  0.256               
## PopulatinVA 0.297  0.263  0.296  0.297  0.297  0.262  0.278        
## PopulatinYE 0.291  0.257  0.290  0.291  0.291  0.255  0.272  0.290

p_DT_A_F_Hoedjes <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Hoedjes')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_Hoedjes", x="Population", y = "DT_EggAdult")+ ylim(c(min_DT_A, max_DT_A))+ droseu_theme

## quartz_off_screen 
##                 2

p_DT_A_F_pop_Hoedjes <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Hoedjes')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_pop_Hoedjes", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ geom_text(data = model_means_DT_A_F_Hoedjes_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

## quartz_off_screen 
##                 2

p_DT_A_M_Hoedjes <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Hoedjes')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_Hoedjes", x="Population", y = "DT_EggAdult")+ ylim(c(min_DT_A, max_DT_A))+ droseu_theme

## quartz_off_screen 
##                 2

p_DT_A_M_pop_Hoedjes <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Hoedjes')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_pop_Hoedjes", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ geom_text(data = model_means_DT_A_M_Hoedjes_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

## quartz_off_screen 
##                 2

Schmidt Lab

DT_A_F_lmer_Schmidt <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population), data = d_DT_A_F[d_DT_A_F$Supervisor.PI == "Schmidt",])
capture.output(summary(DT_A_F_lmer_Schmidt),file = "DevelopmentTime/DT_A_F_lmer_Schmidt_sum.txt")
capture.output(anova(DT_A_F_lmer_Schmidt),file = "DevelopmentTime/DT_A_F_lmer_Schmidt.txt")
capture.output(emmeans(DT_A_F_lmer_Schmidt, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_F_lmer_Schmidt_tk.txt")

model_means_DT_A_F_Schmidt <- emmeans(object = DT_A_F_lmer_Schmidt, specs = ~ Population)
model_means_DT_A_F_Schmidt_cld <- cld(object = model_means_DT_A_F_Schmidt, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_F_lmer_Schmidt)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 4258902  473211     9 144.94  2183.1 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_F_lmer_Schmidt)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population)
##    Data: d_DT_A_F[d_DT_A_F$Supervisor.PI == "Schmidt", ]
## 
## REML criterion at convergence: 14067.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.5685 -0.4672 -0.0841  0.3244  9.1996 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 382.1    19.55   
##  Residual                    216.8    14.72   
## Number of obs: 1664, groups:  Line:Population, 159
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  218.139      4.503 141.010   48.44   <2e-16 ***
## PopulationGI  226.606      6.229 154.876   36.38   <2e-16 ***
## PopulationKA  220.027      4.622 140.746   47.61   <2e-16 ***
## PopulationMA  224.372      4.649 143.674   48.27   <2e-16 ***
## PopulationMU  221.157      4.491 139.463   49.25   <2e-16 ***
## PopulationRE  230.050      5.299 149.276   43.41   <2e-16 ***
## PopulationUM  222.182      5.168 137.621   42.99   <2e-16 ***
## PopulationVA  245.627      4.587 149.719   53.55   <2e-16 ***
## PopulationYE  222.365      4.581 150.071   48.54   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

DT_A_M_lmer_Schmidt <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population), data = d_DT_A_M[d_DT_A_M$Supervisor.PI == "Schmidt",])
capture.output(summary(DT_A_M_lmer_Schmidt),file = "DevelopmentTime/DT_A_M_lmer_Schmidt_sum.txt")
capture.output(anova(DT_A_M_lmer_Schmidt),file = "DevelopmentTime/DT_A_M_lmer_Schmidt.txt")
capture.output(emmeans(DT_A_M_lmer_Schmidt, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_M_lmer_Schmidt_tk.txt")

model_means_DT_A_M_Schmidt <- emmeans(object = DT_A_M_lmer_Schmidt, specs = ~ Population)
model_means_DT_A_M_Schmidt_cld <- cld(object = model_means_DT_A_M_Schmidt, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_M_lmer_Schmidt)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 5393416  599268     9 141.65  2017.7 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_M_lmer_Schmidt)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population)
##    Data: d_DT_A_M[d_DT_A_M$Supervisor.PI == "Schmidt", ]
## 
## REML criterion at convergence: 12831.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.8636 -0.4961 -0.1229  0.3065  7.0181 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 416.9    20.42   
##  Residual                    297.0    17.23   
## Number of obs: 1464, groups:  Line:Population, 157
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  226.013      4.743 135.966   47.65   <2e-16 ***
## PopulationGI  243.117      6.598 150.387   36.85   <2e-16 ***
## PopulationKA  219.913      4.876 136.693   45.10   <2e-16 ***
## PopulationMA  228.741      4.931 141.233   46.39   <2e-16 ***
## PopulationMU  226.716      4.871 136.720   46.54   <2e-16 ***
## PopulationRE  231.651      5.636 149.746   41.10   <2e-16 ***
## PopulationUM  229.105      5.464 134.715   41.93   <2e-16 ***
## PopulationVA  250.048      4.932 142.913   50.70   <2e-16 ***
## PopulationYE  225.938      4.861 148.803   46.48   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_DT_A_F_Schmidt <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_Schmidt", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ droseu_theme

p_DT_A_F_Schmidt

pdf(file="DevelopmentTime/p_DT_A_F_Schmidt.pdf",width=8, height=5)
p_DT_A_F_Schmidt
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_F_pop_Schmidt <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_pop_Schmidt", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+geom_text(data = model_means_DT_A_F_Schmidt_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_DT_A_F_pop_Schmidt

pdf(file="DevelopmentTime/p_DT_A_F_pop_Schmidt.pdf",width=8, height=5)
p_DT_A_F_pop_Schmidt
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_M_Schmidt <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_Schmidt", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ droseu_theme

p_DT_A_M_Schmidt

pdf(file="DevelopmentTime/p_DT_A_M_Schmidt.pdf",width=8, height=5)
p_DT_A_M_Schmidt
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_M_pop_Schmidt <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_pop_Schmidt", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+geom_text(data = model_means_DT_A_M_Schmidt_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

p_DT_A_M_pop_Schmidt

pdf(file="DevelopmentTime/p_DT_A_M_pop_Schmidt.pdf",width=8, height=5)
p_DT_A_M_pop_Schmidt
dev.off()

## quartz_off_screen 
##                 2

Stamenkovic-Radak Lab

DT_A_F_lmer_StamenkovicRadak <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population)+ (1|Batch) + (1|ReplicateVial : Line), data = d_DT_A_F[d_DT_A_F$Supervisor.PI == "StamenkovicRadak",])
capture.output(summary(DT_A_F_lmer_StamenkovicRadak),file = "DevelopmentTime/DT_A_F_lmer_StamenkovicRadak_sum.txt")
capture.output(anova(DT_A_F_lmer_StamenkovicRadak),file = "DevelopmentTime/DT_A_F_lmer_StamenkovicRadak.txt")
capture.output(emmeans(DT_A_F_lmer_StamenkovicRadak, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_F_lmer_StamenkovicRadak_tk.txt")

model_means_DT_A_F_StamenkovicRadak <- emmeans(object = DT_A_F_lmer_StamenkovicRadak, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 6200' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 6200)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 6200' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 6200)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_A_F_StamenkovicRadak_cld <- cld(object = model_means_DT_A_F_StamenkovicRadak, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_F_lmer_StamenkovicRadak)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 1580329  175592     9 10.415  574.67 8.483e-13 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_F_lmer_StamenkovicRadak)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: d_DT_A_F[d_DT_A_F$Supervisor.PI == "StamenkovicRadak", ]
## 
## REML criterion at convergence: 54203.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.5991 -0.5815 -0.1328  0.3761  6.5733 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept) 148.74   12.196  
##  Line:Population    (Intercept) 175.07   13.231  
##  Batch              (Intercept)  40.54    6.367  
##  Residual                       305.55   17.480  
## Number of obs: 6200, groups:  
## ReplicateVial:Line, 494; Line:Population, 165; Batch, 4
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  237.623      4.734  10.934   50.19 2.80e-14 ***
## PopulationGI  245.089      5.318  16.968   46.08  < 2e-16 ***
## PopulationKA  229.151      4.666  10.342   49.12 1.35e-13 ***
## PopulationMA  246.250      4.693  10.558   52.47 4.13e-14 ***
## PopulationMU  237.559      4.666  10.341   50.91 9.37e-14 ***
## PopulationRE  255.347      4.976  13.236   51.32  < 2e-16 ***
## PopulationUM  252.917      4.816  11.692   52.52 3.01e-15 ***
## PopulationVA  251.781      4.750  11.062   53.01 1.15e-14 ***
## PopulationYE  240.080      4.692  10.574   51.17 5.18e-14 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.404                                                  
## PopulatinKA 0.459  0.408                                           
## PopulatinMA 0.459  0.407  0.463                                    
## PopulatinMU 0.460  0.409  0.466  0.465                             
## PopulatinRE 0.430  0.381  0.437  0.435  0.437                      
## PopulatinUM 0.444  0.394  0.451  0.449  0.451  0.424               
## PopulatinVA 0.453  0.405  0.457  0.458  0.459  0.429  0.442        
## PopulatinYE 0.456  0.406  0.463  0.460  0.463  0.434  0.448  0.455

DT_A_M_lmer_StamenkovicRadak <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population) + (1|Batch) + (1|ReplicateVial : Line), data = d_DT_A_M[d_DT_A_M$Supervisor.PI == "StamenkovicRadak",])
capture.output(summary(DT_A_M_lmer_StamenkovicRadak),file = "DevelopmentTime/DT_A_M_lmer_StamenkovicRadak_sum.txt")
capture.output(anova(DT_A_M_lmer_StamenkovicRadak),file = "DevelopmentTime/DT_A_M_lmer_StamenkovicRadak.txt")
capture.output(emmeans(DT_A_M_lmer_StamenkovicRadak, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_M_lmer_StamenkovicRadak_tk.txt")

model_means_DT_A_M_StamenkovicRadak <- emmeans(object = DT_A_M_lmer_StamenkovicRadak, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 5815' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 5815)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 5815' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 5815)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_A_M_StamenkovicRadak_cld <- cld(object = model_means_DT_A_M_StamenkovicRadak, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_M_lmer_StamenkovicRadak)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 1933317  214813     9 10.511     652 3.489e-13 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_M_lmer_StamenkovicRadak)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: d_DT_A_M[d_DT_A_M$Supervisor.PI == "StamenkovicRadak", ]
## 
## REML criterion at convergence: 51282.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.9954 -0.5866 -0.1542  0.3744  6.6404 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept) 152.27   12.340  
##  Line:Population    (Intercept) 162.25   12.738  
##  Batch              (Intercept)  37.14    6.094  
##  Residual                       329.47   18.151  
## Number of obs: 5815, groups:  
## ReplicateVial:Line, 494; Line:Population, 165; Batch, 4
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  242.718      4.584  11.349   52.95 6.00e-15 ***
## PopulationGI  251.724      5.164  17.775   48.75  < 2e-16 ***
## PopulationKA  234.825      4.519  10.752   51.96 2.92e-14 ***
## PopulationMA  250.468      4.538  10.905   55.19 1.07e-14 ***
## PopulationMU  244.124      4.518  10.734   54.03 2.01e-14 ***
## PopulationRE  259.635      4.823  13.793   53.83  < 2e-16 ***
## PopulationUM  257.083      4.678  12.289   54.96 4.46e-16 ***
## PopulationVA  259.384      4.599  11.483   56.40 2.14e-15 ***
## PopulationYE  246.043      4.544  10.989   54.15 1.08e-14 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.394                                                  
## PopulatinKA 0.448  0.398                                           
## PopulatinMA 0.449  0.397  0.453                                    
## PopulatinMU 0.450  0.399  0.455  0.455                             
## PopulatinRE 0.420  0.371  0.426  0.426  0.427                      
## PopulatinUM 0.433  0.383  0.439  0.437  0.439  0.412               
## PopulatinVA 0.443  0.395  0.447  0.448  0.449  0.419  0.431        
## PopulatinYE 0.446  0.396  0.452  0.450  0.452  0.424  0.437  0.444

p_DT_A_F_StamenkovicRadak <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_StamenkovicRadak", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ droseu_theme

p_DT_A_F_StamenkovicRadak

pdf(file="DevelopmentTime/p_DT_A_F_StamenkovicRadak.pdf",width=8, height=5)
p_DT_A_F_StamenkovicRadak
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_F_pop_StamenkovicRadak <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_pop_StamenkovicRadak", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ geom_text(data = model_means_DT_A_F_StamenkovicRadak_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_DT_A_F_pop_StamenkovicRadak

pdf(file="DevelopmentTime/p_DT_A_F_pop_StamenkovicRadak.pdf",width=8, height=5)
p_DT_A_F_pop_StamenkovicRadak
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_M_StamenkovicRadak <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_StamenkovicRadak", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+ droseu_theme

p_DT_A_M_StamenkovicRadak

pdf(file="DevelopmentTime/p_DT_A_M_StamenkovicRadak.pdf",width=8, height=5)
p_DT_A_M_StamenkovicRadak
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_M_pop_StamenkovicRadak <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_pop_StamenkovicRadak", x="Population", y = "DT_EggAdult")+ ylim(c(min_DT_A, max_DT_A))+ geom_text(data = model_means_DT_A_M_StamenkovicRadak_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_DT_A_M_pop_StamenkovicRadak

pdf(file="DevelopmentTime/p_DT_A_M_pop_StamenkovicRadak.pdf",width=8, height=5)
p_DT_A_M_pop_StamenkovicRadak
dev.off()

## quartz_off_screen 
##                 2

Zwaan Lab

DT_A_F_lmer_Zwaan <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population) + (1|Batch) + (1|ReplicateVial : Line), data = d_DT_A_F[d_DT_A_F$Supervisor.PI == "Zwaan",])
capture.output(summary(DT_A_F_lmer_Zwaan),file = "DevelopmentTime/DT_A_F_lmer_Zwaan_sum.txt")
capture.output(anova(DT_A_F_lmer_Zwaan),file = "DevelopmentTime/DT_A_F_lmer_Zwaan.txt")
capture.output(emmeans(DT_A_F_lmer_Zwaan, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_F_lmer_Zwaan_tk.txt")

model_means_DT_A_F_Zwaan <- emmeans(object = DT_A_F_lmer_Zwaan, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 7330' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 7330)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 7330' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 7330)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_A_F_Zwaan_cld <- cld(object = model_means_DT_A_F_Zwaan, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_F_lmer_Zwaan)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 1572099  174678     9     2  1565.5 0.0006385 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_F_lmer_Zwaan)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: d_DT_A_F[d_DT_A_F$Supervisor.PI == "Zwaan", ]
## 
## REML criterion at convergence: 55980.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.4229 -0.5259 -0.1848  0.2790  6.6947 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept)   8.246   2.872  
##  Line:Population    (Intercept)  25.504   5.050  
##  Batch              (Intercept)   6.706   2.590  
##  Residual                       111.578  10.563  
## Number of obs: 7330, groups:  
## ReplicateVial:Line, 521; Line:Population, 169; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  220.748      2.213   2.005   99.74 9.85e-05 ***
## PopulationGI  224.867      2.336   2.485   96.27 1.67e-05 ***
## PopulationKA  223.320      2.216   2.016  100.77 9.25e-05 ***
## PopulationMA  224.220      2.215   2.013  101.21 9.26e-05 ***
## PopulationMU  222.126      2.228   2.060   99.68 7.93e-05 ***
## PopulationRE  226.261      2.301   2.343   98.31 2.73e-05 ***
## PopulationUM  222.349      2.251   2.144   98.79 5.82e-05 ***
## PopulationVA  224.948      2.221   2.032  101.30 8.57e-05 ***
## PopulationYE  221.280      2.248   2.135   98.42 6.09e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.649                                                  
## PopulatinKA 0.684  0.648                                           
## PopulatinMA 0.684  0.649  0.683                                    
## PopulatinMU 0.680  0.645  0.679  0.679                             
## PopulatinRE 0.659  0.626  0.658  0.658  0.654                      
## PopulatinUM 0.673  0.639  0.672  0.673  0.669  0.648               
## PopulatinVA 0.682  0.647  0.681  0.682  0.678  0.657  0.671        
## PopulatinYE 0.674  0.640  0.673  0.674  0.670  0.649  0.663  0.672

DT_A_M_lmer_Zwaan <- lmer(DT_EggAdult ~ -1 + Population + (1|Line:Population) + (1|Batch) + (1|ReplicateVial : Line), data = d_DT_A_M[d_DT_A_M$Supervisor.PI == "Zwaan",])
capture.output(summary(DT_A_M_lmer_Zwaan),file = "DevelopmentTime/DT_A_M_lmer_Zwaan_sum.txt")
capture.output(anova(DT_A_M_lmer_Zwaan),file = "DevelopmentTime/DT_A_M_lmer_Zwaan.txt")
capture.output(emmeans(DT_A_M_lmer_Zwaan, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DevelopmentTime/DT_A_M_lmer_Zwaan_tk.txt")

model_means_DT_A_M_Zwaan <- emmeans(object = DT_A_M_lmer_Zwaan, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 6394' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 6394)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 6394' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 6394)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DT_A_M_Zwaan_cld <- cld(object = model_means_DT_A_M_Zwaan, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DT_A_M_lmer_Zwaan)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF DenDF F value   Pr(>F)    
## Population 1363231  151470     9     2  1235.6 0.000809 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DT_A_M_lmer_Zwaan)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DT_EggAdult ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: d_DT_A_M[d_DT_A_M$Supervisor.PI == "Zwaan", ]
## 
## REML criterion at convergence: 49447.1
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.9922 -0.5259 -0.1942  0.2460  6.2037 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept)   7.491   2.737  
##  Line:Population    (Intercept)  29.635   5.444  
##  Batch              (Intercept)   8.828   2.971  
##  Residual                       122.593  11.072  
## Number of obs: 6394, groups:  
## ReplicateVial:Line, 519; Line:Population, 169; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  225.736      2.488   1.867   90.74 0.000204 ***
## PopulationGI  229.333      2.615   2.278   87.69 4.54e-05 ***
## PopulationKA  226.969      2.485   1.859   91.33 0.000207 ***
## PopulationMA  227.755      2.491   1.876   91.44 0.000194 ***
## PopulationMU  225.545      2.496   1.893   90.35 0.000185 ***
## PopulationRE  229.915      2.583   2.168   89.02 6.66e-05 ***
## PopulationUM  225.065      2.524   1.978   89.17 0.000137 ***
## PopulationVA  227.789      2.495   1.889   91.30 0.000185 ***
## PopulationYE  223.114      2.522   1.973   88.45 0.000142 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.679                                                  
## PopulatinKA 0.714  0.679                                           
## PopulatinMA 0.712  0.678  0.713                                    
## PopulatinMU 0.711  0.676  0.711  0.710                             
## PopulatinRE 0.687  0.655  0.688  0.686  0.685                      
## PopulatinUM 0.703  0.669  0.704  0.702  0.701  0.678               
## PopulatinVA 0.711  0.677  0.712  0.710  0.709  0.685  0.701        
## PopulatinYE 0.703  0.670  0.704  0.703  0.701  0.678  0.694  0.701

p_DT_A_F_Zwaan <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Zwaan')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_Zwaan", x="Population", y = "DT_EggAdult")+ ylim(c(min_DT_A, max_DT_A))+ droseu_theme

p_DT_A_F_Zwaan

pdf(file="DevelopmentTime/p_DT_A_F_Zwaan.pdf",width=8, height=5)
p_DT_A_F_Zwaan
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_F_pop_Zwaan <- ggplot(data = (subset(d_DT_A_F,Supervisor.PI=='Zwaan')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_F_pop_Zwaan", x="Population", y = "DT_EggAdult")+ ylim(c(min_DT_A, max_DT_A))+geom_text(data = model_means_DT_A_F_Zwaan_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

p_DT_A_F_pop_Zwaan

pdf(file="DevelopmentTime/p_DT_A_F_pop_Zwaan.pdf",width=8, height=5)
p_DT_A_F_pop_Zwaan
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_M_Zwaan <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Zwaan')), aes(x=Population_Lat, y=DT_EggAdult, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_Zwaan", x="Population", y = "DT_EggAdult")+ ylim(c(min_DT_A, max_DT_A))+ droseu_theme

p_DT_A_M_Zwaan

pdf(file="DevelopmentTime/p_DT_A_M_Zwaan.pdf",width=8, height=5)
p_DT_A_M_Zwaan
dev.off()

## quartz_off_screen 
##                 2

p_DT_A_M_pop_Zwaan <- ggplot(data = (subset(d_DT_A_M,Supervisor.PI=='Zwaan')), aes(x=Population_Lat, y=DT_EggAdult, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DT_A_M_pop_Zwaan", x="Population", y = "DT_EggAdult") + ylim(c(min_DT_A, max_DT_A))+geom_text(data = model_means_DT_A_M_Zwaan_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

p_DT_A_M_pop_Zwaan

pdf(file="DevelopmentTime/p_DT_A_M_pop_Zwaan.pdf",width=8, height=5)
p_DT_A_M_pop_Zwaan
dev.off

## function (which = dev.cur()) 
## {
##     if (which == 1) 
##         stop("cannot shut down device 1 (the null device)")
##     .External(C_devoff, as.integer(which))
##     dev.cur()
## }
## <bytecode: 0x7f8af198e190>
## <environment: namespace:grDevices>

Lab correlations

intersect_Line_DT_A  <- Reduce(intersect, list(d_DT_A$Line[d_DT_A$Supervisor.PI == "Gibert"],
                                          d_DT_A$Line[d_DT_A$Supervisor.PI == "Grath"],
                                          d_DT_A$Line[d_DT_A$Supervisor.PI == "Hoedjes"],
                                          d_DT_A$Line[d_DT_A$Supervisor.PI == "Schmidt"],
                                          d_DT_A$Line[d_DT_A$Supervisor.PI == "StamenkovicRadak"],
                                          d_DT_A$Line[d_DT_A$Supervisor.PI == "Zwaan"]))
write(intersect_Line_DT_A, file ="DevelopmentTime/intersect_Line_DT_A.txt", sep = "\t")

intersect_Pop_DT_A  <- Reduce(intersect, list(d_DT_A$Population[d_DT_A$Supervisor.PI == "Gibert"],
                                          d_DT_A$Population[d_DT_A$Supervisor.PI == "Grath"],
                                          d_DT_A$Population[d_DT_A$Supervisor.PI == "Hoedjes"],
                                          d_DT_A$Population[d_DT_A$Supervisor.PI == "Schmidt"],
                                          d_DT_A$Population[d_DT_A$Supervisor.PI == "StamenkovicRadak"],
                                          d_DT_A$Population[d_DT_A$Supervisor.PI == "Zwaan"]))
write(intersect_Pop_DT_A, file ="DevelopmentTime/intersect_Pop_DT_A.txt", sep = "\t")

There are 3 populations and 22 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

DT_A_fitted <- combineFitted(labs = rep(c("Gibert", "Grath", "Hoedjes", "Schmidt", "StamenkovicRadak", "Zwaan"), 2), sex = rep(c("F", "M"), each = 6), models = list(DT_A_F_lmer_Gibert, DT_A_F_lmer_Grath, DT_A_F_lmer_Hoedjes, DT_A_F_lmer_Schmidt, DT_A_F_lmer_StamenkovicRadak, DT_A_F_lmer_Zwaan, DT_A_M_lmer_Gibert, DT_A_M_lmer_Grath, DT_A_M_lmer_Hoedjes, DT_A_M_lmer_Schmidt, DT_A_M_lmer_StamenkovicRadak, DT_A_M_lmer_Zwaan))

p_DT_A_F_lab_cor <- function() scatterPlotMatrix(DT_A_fitted, sex = "F")
p_DT_A_M_lab_cor <- function() scatterPlotMatrix(DT_A_fitted, sex = "M")

p_DT_A_F_lab_cor()

pdf(file="DevelopmentTime/p_DT_A_F_lab_cor.pdf",width=8, height=8)
p_DT_A_F_lab_cor()
invisible(dev.off())

p_DT_A_M_lab_cor()

pdf(file="DevelopmentTime/p_DT_A_M_lab_cor.pdf",width=8, height=8)
p_DT_A_M_lab_cor()
invisible(dev.off())

Dry weight

dir.create(file.path("DryWeight"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Colinet Lab : Sapho-Lou Marti , Hervé Colinet

Hoedjes Lab : Hristina Kostic, Katja Hoedjes

Onder Lab : Seda Coskun, Senel Selin Senkal, Dogus Can, Banu Sebnem Onder

Data Reformatting

Read data in R

d_DW <- read.csv("MasterSheets_Oct21_git/DW_MasterSheet_Oct21.csv")

str(d_DW)

## 'data.frame':    24161 obs. of  12 variables:
##  $ Supervisor.PI: chr  "Colinet" "Colinet" "Colinet" "Colinet" ...
##  $ Diet         : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch        : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population   : chr  "AK" "AK" "AK" "AK" ...
##  $ Line         : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Sex          : chr  "F" "F" "F" "F" ...
##  $ Individual   : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ DW_micrograms: num  0.394 0.473 0.581 0.594 0.454 0.486 0.581 0.584 0.596 0.573 ...
##  $ Country      : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude     : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude    : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude     : int  88 88 88 88 88 88 88 88 88 88 ...

d_DW$Supervisor.PI <- as.factor(d_DW$Supervisor.PI)
d_DW$Diet <- as.factor(d_DW$Diet)
d_DW$Batch <- as.factor(d_DW$Batch)
d_DW$Population <- as.factor(d_DW$Population)
d_DW$Population_Lat <- factor(d_DW$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_DW$Population_Lon <- factor(d_DW$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_DW$Population_Alt <- factor(d_DW$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_DW$Line <- as.factor(d_DW$Line)
d_DW$Sex <- as.factor(d_DW$Sex)
d_DW$DW_micrograms <- as.numeric(d_DW$DW_micrograms)

str(d_DW)

## 'data.frame':    24161 obs. of  15 variables:
##  $ Supervisor.PI : Factor w/ 3 levels "Colinet","Hoedjes",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet          : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch         : Factor w/ 5 levels "1","2","3","4",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population    : Factor w/ 9 levels "AK","GI","KA",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Line          : Factor w/ 170 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex           : Factor w/ 2 levels "F","M": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Individual    : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ DW_micrograms : num  0.394 0.473 0.581 0.594 0.454 0.486 0.581 0.584 0.596 0.573 ...
##  $ Country       : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude      : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude     : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude      : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat: Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon: Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt: Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_DW_F <-subset(d_DW,Sex=='F')
d_DW_M <-subset(d_DW,Sex=='M')

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

Descriptive statistics at the line level, without batch information :

Descriptive statistics at the population level, with batch information :

Descriptive statistics at the population level, without batch information :

Plots and Linear Models by Lab

Overview

min_DW <- min(d_DW$DW_micrograms)
max_DW <- max(d_DW$DW_micrograms)

y-axis is scaled by the minimum (0.136) and maximum (0.73) values in the full data set.

pdf(file="DryWeight/p_DW_F_all.pdf",width=8, height=5)
p_DW_F_all <- ggplot(data = d_DW_F, aes(x=Population_Lat, y=DW_micrograms, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_F_all", x="Population", y = "DW_micrograms")

p_DW_F_all + ylim(c(min_DW, max_DW))+ droseu_theme
dev.off()

## quartz_off_screen 
##                 2

pdf(file="DryWeight/p_DW_M_all.pdf",width=8, height=5)
p_DW_M_all <- ggplot(data = d_DW_M, aes(x=Population_Lat, y=DW_micrograms, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_M_all", x="Population", y = "DW_micrograms")

p_DW_M_all + ylim(c(min_DW, max_DW))+ droseu_theme
dev.off()

## quartz_off_screen 
##                 2

Colinet Lab

DW_F_lmer_Colinet <- lmer(DW_micrograms ~ -1 + Population + (1|Population:Line) + (1|Batch), data = d_DW_F[d_DW_F$Supervisor.PI == "Colinet",])
capture.output(summary(DW_F_lmer_Colinet),file = "DryWeight/DW_F_lmer_Colinet_sum.txt")
capture.output(anova(DW_F_lmer_Colinet),file = "DryWeight/DW_F_lmer_Colinet.txt")
capture.output(emmeans(DW_F_lmer_Colinet, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DryWeight/DW_F_lmer_Colinet_tk.txt")

model_means_DW_F_Colinet <- emmeans(object = DW_F_lmer_Colinet, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 3982' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 3982)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 3982' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 3982)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DW_F_Colinet_cld <- cld(object = model_means_DW_F_Colinet, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DW_F_lmer_Colinet)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value   Pr(>F)   
## Population 7.8449 0.87166     9     2  323.43 0.003086 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DW_F_lmer_Colinet)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DW_micrograms ~ -1 + Population + (1 | Population:Line) + (1 |  
##     Batch)
##    Data: d_DW_F[d_DW_F$Supervisor.PI == "Colinet", ]
## 
## REML criterion at convergence: -11771.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.2178 -0.6128  0.0273  0.6249  4.1344 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Population:Line (Intercept) 0.001328 0.03645 
##  Batch           (Intercept) 0.000219 0.01480 
##  Residual                    0.002695 0.05191 
## Number of obs: 3982, groups:  Population:Line, 166; Batch, 3
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  0.46881    0.01205 5.84387   38.89 2.79e-08 ***
## PopulationGI  0.47949    0.01328 8.25191   36.11 2.23e-10 ***
## PopulationKA  0.49849    0.01239 6.27649   40.23 8.20e-09 ***
## PopulationMA  0.49998    0.01220 6.13762   40.98 1.02e-08 ***
## PopulationMU  0.48458    0.01205 5.84387   40.20 2.30e-08 ***
## PopulationRE  0.51478    0.01328 8.24981   38.77 1.25e-10 ***
## PopulationUM  0.50102    0.01287 7.28302   38.94 1.01e-09 ***
## PopulationVA  0.49685    0.01239 6.27649   40.09 8.37e-09 ***
## PopulationYE  0.48099    0.01205 5.84480   39.90 2.40e-08 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.447                                                  
## PopulatinKA 0.479  0.490                                           
## PopulatinMA 0.497  0.446  0.478                                    
## PopulatinMU 0.504  0.447  0.479  0.497                             
## PopulatinRE 0.447  0.455  0.490  0.446  0.447                      
## PopulatinUM 0.462  0.471  0.509  0.460  0.462  0.471               
## PopulatinVA 0.479  0.490  0.531  0.478  0.479  0.490  0.509        
## PopulatinYE 0.504  0.447  0.479  0.497  0.504  0.447  0.462  0.479

DW_M_lmer_Colinet <- lmer(DW_micrograms ~ -1 + Population + (1|Population:Line) + (1|Batch), data = d_DW_M[d_DW_M$Supervisor.PI == "Colinet",])

## boundary (singular) fit: see ?isSingular

capture.output(summary(DW_M_lmer_Colinet),file = "DryWeight/DW_M_lmer_Colinet_sum.txt")
capture.output(anova(DW_M_lmer_Colinet),file = "DryWeight/DW_M_lmer_Colinet.txt")
capture.output(emmeans(DW_M_lmer_Colinet, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DryWeight/DW_M_lmer_Colinet_tk.txt")

model_means_DW_M_Colinet <- emmeans(object = DW_M_lmer_Colinet, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 3975' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 3975)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 3975' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 3975)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DW_M_Colinet_cld <- cld(object = model_means_DW_M_Colinet, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DW_M_lmer_Colinet)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 22.773  2.5304     9 157.01  3533.8 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DW_M_lmer_Colinet)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DW_micrograms ~ -1 + Population + (1 | Population:Line) + (1 |  
##     Batch)
##    Data: d_DW_M[d_DW_M$Supervisor.PI == "Colinet", ]
## 
## REML criterion at convergence: -17014.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.6407 -0.5828 -0.0246  0.5790  7.4085 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev.
##  Population:Line (Intercept) 0.0003446 0.01856 
##  Batch           (Intercept) 0.0000000 0.00000 
##  Residual                    0.0007161 0.02676 
## Number of obs: 3975, groups:  Population:Line, 166; Batch, 3
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## PopulationAK 2.595e-01  4.327e-03 1.570e+02   59.97   <2e-16 ***
## PopulationGI 2.604e-01  4.999e-03 1.574e+02   52.09   <2e-16 ***
## PopulationKA 2.791e-01  4.327e-03 1.570e+02   64.51   <2e-16 ***
## PopulationMA 2.711e-01  4.439e-03 1.569e+02   61.08   <2e-16 ***
## PopulationMU 2.721e-01  4.327e-03 1.569e+02   62.89   <2e-16 ***
## PopulationRE 2.761e-01  4.996e-03 1.569e+02   55.26   <2e-16 ***
## PopulationUM 2.758e-01  4.693e-03 1.569e+02   58.78   <2e-16 ***
## PopulationVA 2.675e-01  4.327e-03 1.570e+02   61.83   <2e-16 ***
## PopulationYE 2.492e-01  4.327e-03 1.570e+02   57.59   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

p_DW_F_Colinet <- ggplot(data = (subset(d_DW_F,Supervisor.PI=='Colinet')), aes(x=Population_Lat, y=DW_micrograms, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_F_Colinet", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+ droseu_theme

## quartz_off_screen 
##                 2

p_DW_F_pop_Colinet <- ggplot(data = (subset(d_DW_F,Supervisor.PI=='Colinet')), aes(x=Population_Lat, y=DW_micrograms, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_F_pop_Colinet", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+geom_text(data = model_means_DW_F_Colinet_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

## quartz_off_screen 
##                 2

p_DW_M_Colinet <- ggplot(data = (subset(d_DW_M,Supervisor.PI=='Colinet')), aes(x=Population_Lat, y=DW_micrograms, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_M_Colinet", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+ droseu_theme

## quartz_off_screen 
##                 2

p_DW_M_pop_Colinet <- ggplot(data = (subset(d_DW_M,Supervisor.PI=='Colinet')), aes(x=Population_Lat, y=DW_micrograms, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_M_pop_Colinet", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+geom_text(data = model_means_DW_M_Colinet_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

## quartz_off_screen 
##                 2

Hoedjes Lab

DW_F_lmer_Hoedjes <- lmer(DW_micrograms ~ -1 + Population + (1|Population:Line) + (1|Batch), data = d_DW_F[d_DW_F$Supervisor.PI == "Hoedjes",])
capture.output(summary(DW_F_lmer_Hoedjes),file = "DryWeight/DW_F_lmer_Hoedjes_sum.txt")
capture.output(anova(DW_F_lmer_Hoedjes),file = "DryWeight/DW_F_lmer_Hoedjes.txt")
capture.output(emmeans(DW_F_lmer_Hoedjes, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DryWeight/DW_F_lmer_Hoedjes_tk.txt")

model_means_DW_F_Hoedjes <- emmeans(object = DW_F_lmer_Hoedjes, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 4004' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 4004)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 4004' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 4004)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DW_F_Hoedjes_cld <- cld(object = model_means_DW_F_Hoedjes, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DW_F_lmer_Hoedjes)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 16.301  1.8112     9 19.473   804.2 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DW_F_lmer_Hoedjes)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DW_micrograms ~ -1 + Population + (1 | Population:Line) + (1 |  
##     Batch)
##    Data: d_DW_F[d_DW_F$Supervisor.PI == "Hoedjes", ]
## 
## REML criterion at convergence: -12532.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.6857 -0.6210 -0.0043  0.6319  3.4401 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev.
##  Population:Line (Intercept) 1.272e-03 0.035659
##  Batch           (Intercept) 8.069e-05 0.008983
##  Residual                    2.252e-03 0.047457
## Number of obs: 4004, groups:  Population:Line, 167; Batch, 4
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## PopulationAK  0.437373   0.009404 34.224775   46.51   <2e-16 ***
## PopulationGI  0.448305   0.010572 48.980304   42.41   <2e-16 ***
## PopulationKA  0.458818   0.009405 34.225455   48.78   <2e-16 ***
## PopulationMA  0.443136   0.009406 34.241181   47.11   <2e-16 ***
## PopulationMU  0.456749   0.009402 34.185580   48.58   <2e-16 ***
## PopulationRE  0.465470   0.010563 49.059236   44.06   <2e-16 ***
## PopulationUM  0.449962   0.010034 42.068161   44.84   <2e-16 ***
## PopulationVA  0.473101   0.009408 34.269319   50.29   <2e-16 ***
## PopulationYE  0.442165   0.009426 34.396102   46.91   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.203                                                  
## PopulatinKA 0.228  0.202                                           
## PopulatinMA 0.228  0.203  0.228                                    
## PopulatinMU 0.228  0.203  0.228  0.228                             
## PopulatinRE 0.203  0.182  0.203  0.203  0.203                      
## PopulatinUM 0.214  0.191  0.214  0.214  0.213  0.193               
## PopulatinVA 0.228  0.203  0.228  0.228  0.228  0.203  0.214        
## PopulatinYE 0.228  0.202  0.228  0.228  0.228  0.200  0.212  0.228

DW_M_lmer_Hoedjes <- lmer(DW_micrograms ~ -1 + Population + (1|Population:Line) + (1|Batch), data = d_DW_M[d_DW_M$Supervisor.PI == "Hoedjes",])
capture.output(summary(DW_M_lmer_Hoedjes),file = "DryWeight/DW_M_lmer_Hoedjes_sum.txt")
capture.output(anova(DW_M_lmer_Hoedjes),file = "DryWeight/DW_M_lmer_Hoedjes.txt")
capture.output(emmeans(DW_M_lmer_Hoedjes, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DryWeight/DW_M_lmer_Hoedjes_tk.txt")

model_means_DW_M_Hoedjes <- emmeans(object = DW_M_lmer_Hoedjes, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 3997' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 3997)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 3997' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 3997)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DW_M_Hoedjes_cld <- cld(object = model_means_DW_M_Hoedjes, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DW_M_lmer_Hoedjes)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2.2285 0.24762     9 21.575  367.04 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DW_M_lmer_Hoedjes)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DW_micrograms ~ -1 + Population + (1 | Population:Line) + (1 |  
##     Batch)
##    Data: d_DW_M[d_DW_M$Supervisor.PI == "Hoedjes", ]
## 
## REML criterion at convergence: -17309.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.1189 -0.6261 -0.0298  0.6056  5.1645 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev.
##  Population:Line (Intercept) 3.883e-04 0.019704
##  Batch           (Intercept) 9.299e-05 0.009643
##  Residual                    6.746e-04 0.025974
## Number of obs: 3997, groups:  Population:Line, 167; Batch, 5
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## PopulationAK  0.263655   0.006342 14.534898   41.58   <2e-16 ***
## PopulationGI  0.265886   0.006861 19.489321   38.76   <2e-16 ***
## PopulationKA  0.274427   0.006340 14.522154   43.28   <2e-16 ***
## PopulationMA  0.266815   0.006330 14.457218   42.15   <2e-16 ***
## PopulationMU  0.269200   0.006342 14.534547   42.45   <2e-16 ***
## PopulationRE  0.269957   0.006885 19.684454   39.21   <2e-16 ***
## PopulationUM  0.265389   0.006631 17.145310   40.02   <2e-16 ***
## PopulationVA  0.272305   0.006320 14.402207   43.09   <2e-16 ***
## PopulationYE  0.264695   0.006334 14.494587   41.79   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.442                                                  
## PopulatinKA 0.482  0.443                                           
## PopulatinMA 0.481  0.442  0.481                                    
## PopulatinMU 0.482  0.443  0.482  0.481                             
## PopulatinRE 0.443  0.410  0.444  0.444  0.444                      
## PopulatinUM 0.461  0.424  0.461  0.461  0.461  0.426               
## PopulatinVA 0.479  0.441  0.480  0.479  0.480  0.443  0.459        
## PopulatinYE 0.481  0.442  0.481  0.480  0.481  0.443  0.459  0.479

p_DW_F_Hoedjes <- ggplot(data = (subset(d_DW_F,Supervisor.PI=='Hoedjes')), aes(x=Population_Lat, y=DW_micrograms, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_F_Hoedjes", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+ droseu_theme

## quartz_off_screen 
##                 2

p_DW_F_pop_Hoedjes <- ggplot(data = (subset(d_DW_F,Supervisor.PI=='Hoedjes')), aes(x=Population_Lat, y=DW_micrograms, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_F_pop_Hoedjes", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+geom_text(data = model_means_DW_F_Hoedjes_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

## quartz_off_screen 
##                 2

p_DW_M_Hoedjes <- ggplot(data = (subset(d_DW_M,Supervisor.PI=='Hoedjes')), aes(x=Population_Lat, y=DW_micrograms, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_M_Hoedjes", x="Population", y = "DW_micrograms")+ ylim(c(min_DW, max_DW))+ droseu_theme

## quartz_off_screen 
##                 2

p_DW_M_pop_Hoedjes <- ggplot(data = (subset(d_DW_M,Supervisor.PI=='Hoedjes')), aes(x=Population_Lat, y=DW_micrograms, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_M_pop_Hoedjes", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+geom_text(data = model_means_DW_F_Hoedjes_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

## quartz_off_screen 
##                 2

Onder Lab

DW_F_lmer_Onder <- lmer(DW_micrograms ~ -1 + Population + (1|Population:Line) + (1|Batch), data = d_DW_F[d_DW_F$Supervisor.PI == "Onder",])
capture.output(summary(DW_F_lmer_Onder),file = "DryWeight/DW_F_lmer_Onder_sum.txt")
capture.output(anova(DW_F_lmer_Onder),file = "DryWeight/DW_F_lmer_Onder.txt")
capture.output(emmeans(DW_F_lmer_Onder, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DryWeight/DW_F_lmer_Onder_tk.txt")

model_means_DW_F_Onder <- emmeans(object = DW_F_lmer_Onder, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 4102' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 4102)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 4102' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 4102)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DW_F_Onder_cld <- cld(object = model_means_DW_F_Onder, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DW_F_lmer_Onder)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 10.529  1.1699     9 22.027  442.39 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DW_F_lmer_Onder)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DW_micrograms ~ -1 + Population + (1 | Population:Line) + (1 |  
##     Batch)
##    Data: d_DW_F[d_DW_F$Supervisor.PI == "Onder", ]
## 
## REML criterion at convergence: -12234.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.7744 -0.5882  0.0568  0.6380  3.6377 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev.
##  Population:Line (Intercept) 0.0011087 0.03330 
##  Batch           (Intercept) 0.0002207 0.01485 
##  Residual                    0.0026444 0.05142 
## Number of obs: 4102, groups:  Population:Line, 166; Batch, 5
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.42538    0.01032 17.78538   41.21   <2e-16 ***
## PopulationGI  0.43876    0.01129 24.51089   38.85   <2e-16 ***
## PopulationKA  0.46032    0.01033 17.79119   44.57   <2e-16 ***
## PopulationMA  0.46967    0.01032 17.78538   45.50   <2e-16 ***
## PopulationMU  0.44049    0.01028 17.55316   42.86   <2e-16 ***
## PopulationRE  0.46978    0.01106 22.77012   42.47   <2e-16 ***
## PopulationUM  0.46345    0.01082 21.12746   42.83   <2e-16 ***
## PopulationVA  0.47331    0.01025 17.46654   46.16   <2e-16 ***
## PopulationYE  0.42917    0.01062 19.78934   40.40   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.381                                                  
## PopulatinKA 0.409  0.391                                           
## PopulatinMA 0.428  0.381  0.409                                    
## PopulatinMU 0.426  0.383  0.414  0.426                             
## PopulatinRE 0.390  0.368  0.399  0.390  0.393                      
## PopulatinUM 0.388  0.366  0.405  0.388  0.394  0.377               
## PopulatinVA 0.422  0.384  0.416  0.422  0.422  0.392  0.396        
## PopulatinYE 0.412  0.370  0.400  0.412  0.412  0.380  0.382  0.409

DW_M_lmer_Onder <- lmer(DW_micrograms ~ -1 + Population + (1|Population:Line) + (1|Batch), data = d_DW_M[d_DW_M$Supervisor.PI == "Onder",])
capture.output(summary(DW_M_lmer_Onder),file = "DryWeight/DW_M_lmer_Onder_sum.txt")
capture.output(anova(DW_M_lmer_Onder),file = "DryWeight/DW_M_lmer_Onder.txt")
capture.output(emmeans(DW_M_lmer_Onder, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "DryWeight/DW_M_lmer_Onder_tk.txt")

model_means_DW_M_Onder <- emmeans(object = DW_M_lmer_Onder, specs = ~ Population)

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'pbkrtest.limit = 4101' (or larger)
## [or, globally, 'set emm_options(pbkrtest.limit = 4101)' or larger];
## but be warned that this may result in large computation time and memory use.

## Note: D.f. calculations have been disabled because the number of observations exceeds 3000.
## To enable adjustments, add the argument 'lmerTest.limit = 4101' (or larger)
## [or, globally, 'set emm_options(lmerTest.limit = 4101)' or larger];
## but be warned that this may result in large computation time and memory use.

model_means_DW_M_Onder_cld <- cld(object = model_means_DW_M_Onder, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(DW_M_lmer_Onder)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2.9715 0.33016     9 20.883  362.71 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(DW_M_lmer_Onder)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: DW_micrograms ~ -1 + Population + (1 | Population:Line) + (1 |  
##     Batch)
##    Data: d_DW_M[d_DW_M$Supervisor.PI == "Onder", ]
## 
## REML criterion at convergence: -16576.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.2689 -0.6120 -0.0072  0.6481  4.0118 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev.
##  Population:Line (Intercept) 4.358e-04 0.020877
##  Batch           (Intercept) 9.034e-05 0.009505
##  Residual                    9.103e-04 0.030171
## Number of obs: 4101, groups:  Population:Line, 166; Batch, 5
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## PopulationAK  0.255121   0.006505 16.361135   39.22   <2e-16 ***
## PopulationGI  0.251169   0.007103 22.374548   35.36   <2e-16 ***
## PopulationKA  0.263027   0.006508 16.367043   40.42   <2e-16 ***
## PopulationMA  0.271704   0.006505 16.357989   41.77   <2e-16 ***
## PopulationMU  0.261928   0.006478 16.156181   40.43   <2e-16 ***
## PopulationRE  0.264866   0.006961 20.830146   38.05   <2e-16 ***
## PopulationUM  0.267882   0.006813 19.358599   39.32   <2e-16 ***
## PopulationVA  0.266348   0.006465 16.073294   41.20   <2e-16 ***
## PopulationYE  0.252248   0.006691 18.153630   37.70   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.393                                                  
## PopulatinKA 0.422  0.403                                           
## PopulatinMA 0.440  0.393  0.422                                    
## PopulatinMU 0.438  0.396  0.427  0.438                             
## PopulatinRE 0.402  0.380  0.411  0.402  0.405                      
## PopulatinUM 0.401  0.378  0.417  0.401  0.407  0.389               
## PopulatinVA 0.434  0.396  0.429  0.434  0.435  0.404  0.408        
## PopulatinYE 0.425  0.382  0.412  0.425  0.424  0.392  0.394  0.421

p_DW_F_Onder <- ggplot(data = (subset(d_DW_F,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=DW_micrograms, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_F_Onder", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+ droseu_theme

## quartz_off_screen 
##                 2

p_DW_F_pop_Onder <- ggplot(data = (subset(d_DW_F,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=DW_micrograms, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_F_pop_Onder", x="Population", y = "DW_micrograms") + ylim(c(min_DW, max_DW))+geom_text(data = model_means_DW_F_Onder_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

## quartz_off_screen 
##                 2

p_DW_M_Onder <- ggplot(data = (subset(d_DW_M,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=DW_micrograms, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_M_Onder", x="Population", y = "DW_micrograms")+ ylim(c(min_DW, max_DW))+ droseu_theme

## quartz_off_screen 
##                 2

p_DW_M_pop_Onder <- ggplot(data = (subset(d_DW_M,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=DW_micrograms, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_DW_M_pop_Onder", x="Population", y = "DW_micrograms")+ ylim(c(min_DW, max_DW))+geom_text(data = model_means_DW_M_Onder_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

## quartz_off_screen 
##                 2

Lab correlations

intersect_Line_DW  <- Reduce(intersect, list(d_DW$Line[d_DW$Supervisor.PI == "Colinet"],
                                        d_DW$Line[d_DW$Supervisor.PI == "Hoedjes"],
                                        d_DW$Line[d_DW$Supervisor.PI == "Onder"]))

write(intersect_Line_DW, file ="DryWeight/intersect_Line_DW.txt", sep = "\t")

intersect_Pop_DW  <- Reduce(intersect, list(d_DW$Population[d_DW$Supervisor.PI == "Colinet"],
                                        d_DW$Population[d_DW$Supervisor.PI == "Hoedjes"],
                                        d_DW$Population[d_DW$Supervisor.PI == "Onder"]))

write(intersect_Pop_DW, file ="DryWeight/intersect_Pop_DW.txt", sep = "\t")

There are 161 populations and 9 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

DW_fitted <- combineFitted(labs = rep(c("Colinet", "Onder", "Hoedjes"), 2), sex = rep(c("F", "M"), each = 3), models = list(DW_F_lmer_Colinet, DW_F_lmer_Hoedjes, DW_F_lmer_Onder, DW_M_lmer_Colinet, DW_M_lmer_Hoedjes, DW_M_lmer_Onder))

p_DW_F_lab_cor <- function() scatterPlotMatrix(DW_fitted, sex = "F")
p_DW_M_lab_cor <- function() scatterPlotMatrix(DW_fitted, sex = "M")

p_DW_F_lab_cor()

pdf(file="DryWeight/p_DW_F_lab_cor.pdf",width=8, height=8)
p_DW_F_lab_cor()
invisible(dev.off())

p_DW_M_lab_cor()

pdf(file="DryWeight/p_DW_M_lab_cor.pdf",width=8, height=8)
p_DW_M_lab_cor()
invisible(dev.off())

Thorax length

dir.create(file.path("ThoraxLength"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Kozeretska Lab : Svitlana Serga, Alexandra Protsenko, Oleksandr Maistrenko, Iryna Kozeretska

Posnien Lab : Micael Reis, Lennart Hüper

Ritchie Lab : Megan Mcgunnigle, Nicola Cook, Teresa Abaurrea, Michael Ritchie

Schmidt Lab : Amy Goldfischer, Paul Schmidt

Data Reformatting

Read data in R

d_TL <- read.csv("MasterSheets_Oct21_git/TL_MasterSheet_Feb22.csv")
str(d_TL)

## 'data.frame':    13754 obs. of  14 variables:
##  $ Supervisor.PI   : chr  "Ritchie" "Ritchie" "Ritchie" "Ritchie" ...
##  $ Diet            : chr  "S" "S" "S" "S" ...
##  $ Batch           : int  15 15 1 1 1 2 1 1 1 1 ...
##  $ Population      : chr  "GI" "GI" "KA" "KA" ...
##  $ Line            : chr  "GI5" "GI5" "KA1" "KA1" ...
##  $ Sex             : chr  "M" "M" "M" "M" ...
##  $ ReplicateVialOld: int  3 1 3 2 2 2 2 2 2 2 ...
##  $ ReplicateVial   : chr  "Ritchie_15_GI5_3" "Ritchie_15_GI5_1" "Ritchie_1_KA1_3" "Ritchie_1_KA1_2" ...
##  $ Individual      : int  1117 1060 1210 1170 1169 51 1182 1173 1172 1171 ...
##  $ TL_micrometers  : num  300 340 460 514 515 521 552 558 565 580 ...
##  $ Country         : chr  "Spain" "Spain" "Denmark" "Denmark" ...
##  $ Latitude        : num  41.6 41.6 55.9 55.9 55.9 ...
##  $ Longitude       : num  0.62 0.62 10.21 10.21 10.21 ...
##  $ Altitude        : int  173 173 15 15 15 173 15 15 15 15 ...

str(d_TL)

## 'data.frame':    13754 obs. of  14 variables:
##  $ Supervisor.PI   : chr  "Ritchie" "Ritchie" "Ritchie" "Ritchie" ...
##  $ Diet            : chr  "S" "S" "S" "S" ...
##  $ Batch           : int  15 15 1 1 1 2 1 1 1 1 ...
##  $ Population      : chr  "GI" "GI" "KA" "KA" ...
##  $ Line            : chr  "GI5" "GI5" "KA1" "KA1" ...
##  $ Sex             : chr  "M" "M" "M" "M" ...
##  $ ReplicateVialOld: int  3 1 3 2 2 2 2 2 2 2 ...
##  $ ReplicateVial   : chr  "Ritchie_15_GI5_3" "Ritchie_15_GI5_1" "Ritchie_1_KA1_3" "Ritchie_1_KA1_2" ...
##  $ Individual      : int  1117 1060 1210 1170 1169 51 1182 1173 1172 1171 ...
##  $ TL_micrometers  : num  300 340 460 514 515 521 552 558 565 580 ...
##  $ Country         : chr  "Spain" "Spain" "Denmark" "Denmark" ...
##  $ Latitude        : num  41.6 41.6 55.9 55.9 55.9 ...
##  $ Longitude       : num  0.62 0.62 10.21 10.21 10.21 ...
##  $ Altitude        : int  173 173 15 15 15 173 15 15 15 15 ...

d_TL$Supervisor.PI <- as.factor(d_TL$Supervisor.PI)
d_TL$Diet <- as.factor(d_TL$Diet)
d_TL$Batch <- as.factor(d_TL$Batch)
d_TL$Population_Lat <- factor(d_TL$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_TL$Population_Lon <- factor(d_TL$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_TL$Population_Alt <- factor(d_TL$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_TL$Line <- as.factor(d_TL$Line)
d_TL$Sex <- as.factor(d_TL$Sex)
d_TL$ReplicateVial <- as.factor(d_TL$ReplicateVial)
d_TL$TL_micrometers <- as.numeric(d_TL$TL_micrometers)

str(d_TL)

## 'data.frame':    13754 obs. of  17 variables:
##  $ Supervisor.PI   : Factor w/ 4 levels "Kozeretska","Posnien",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ Diet            : Factor w/ 2 levels "NS","S": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Batch           : Factor w/ 12 levels "1","2","3","4",..: 12 12 1 1 1 2 1 1 1 1 ...
##  $ Population      : chr  "GI" "GI" "KA" "KA" ...
##  $ Line            : Factor w/ 166 levels "AK1","AK10","AK11",..: 31 31 36 36 36 21 36 36 36 36 ...
##  $ Sex             : Factor w/ 2 levels "F","M": 2 2 2 2 2 2 2 2 2 2 ...
##  $ ReplicateVialOld: int  3 1 3 2 2 2 2 2 2 2 ...
##  $ ReplicateVial   : Factor w/ 503 levels "Kozeretska_1_AK1_1",..: 298 296 271 270 270 306 270 270 270 270 ...
##  $ Individual      : int  1117 1060 1210 1170 1169 51 1182 1173 1172 1171 ...
##  $ TL_micrometers  : num  300 340 460 514 515 521 552 558 565 580 ...
##  $ Country         : chr  "Spain" "Spain" "Denmark" "Denmark" ...
##  $ Latitude        : num  41.6 41.6 55.9 55.9 55.9 ...
##  $ Longitude       : num  0.62 0.62 10.21 10.21 10.21 ...
##  $ Altitude        : int  173 173 15 15 15 173 15 15 15 15 ...
##  $ Population_Lat  : Factor w/ 9 levels "YE","RE","GI",..: 3 3 7 7 7 3 7 7 7 7 ...
##  $ Population_Lon  : Factor w/ 9 levels "RE","GI","KA",..: 2 2 3 3 3 2 3 3 3 3 ...
##  $ Population_Alt  : Factor w/ 9 levels "KA","AK","GI",..: 3 3 1 1 1 3 1 1 1 1 ...

d_TL_F <-subset(d_TL,Sex=='F')
d_TL_M <-subset(d_TL,Sex=='M')

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

Descriptive statistics at the line level, without batch information :

Descriptive statistics at the population level, with batch information :

Descriptive statistics at the population level, without batch information :

Plots and Linear Models by Lab

Overview

min_TL <- min(d_TL$TL_micrometers)
max_TL <- max(d_TL$TL_micrometers)

y-axis is scaled by the minimum (300) and maximum (1232) values in the full data set.

## quartz_off_screen 
##                 2

p_TL_M <- ggplot(data = d_TL_M, aes(x=Population_Lat, y=TL_micrometers, fill=Supervisor.PI)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_M", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

Kozeretska Lab

TL_F_lmer_Kozeretska <- lmer(TL_micrometers ~ -1 + Population + (1|Line:Population) +
                               (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_TL_F,Supervisor.PI=='Kozeretska')))
capture.output(summary(TL_F_lmer_Kozeretska),file = "ThoraxLength/TL_F_Kozeretska_sum.txt")
capture.output(anova(TL_F_lmer_Kozeretska),file = "ThoraxLength/TL_F_lmer_Kozeretska.txt")
capture.output(emmeans(TL_F_lmer_Kozeretska, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ThoraxLength/TL_F_tk.txt")

anova(TL_F_lmer_Kozeretska)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 101626895 11291877     9 18.383  5376.7 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(TL_F_lmer_Kozeretska)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TL_micrometers ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_TL_F, Supervisor.PI == "Kozeretska"))
## 
## REML criterion at convergence: 51045.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.4816 -0.6029  0.0635  0.6779  3.5476 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  948.79  30.802  
##  Line:Population               (Intercept)  768.84  27.728  
##  Batch                         (Intercept)   16.12   4.015  
##  Residual                                  2100.14  45.827  
## Number of obs: 4810, groups:  
## ReplicateVial:Line:Population, 241; Line:Population, 130; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  936.544      8.986  32.411  104.22   <2e-16 ***
## PopulationGI  935.037     12.728  75.478   73.46   <2e-16 ***
## PopulationKA  911.623     10.772  51.655   84.63   <2e-16 ***
## PopulationMA  964.296      9.484  38.144  101.68   <2e-16 ***
## PopulationMU  942.891      9.327  35.982  101.10   <2e-16 ***
## PopulationRE  962.396     11.023  56.137   87.31   <2e-16 ***
## PopulationUM  941.462     10.896  53.786   86.41   <2e-16 ***
## PopulationVA  958.365      8.796  30.030  108.95   <2e-16 ***
## PopulationYE  950.414     10.721  52.965   88.65   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.071                                                  
## PopulatinKA 0.083  0.059                                           
## PopulatinMA 0.096  0.068  0.079                                    
## PopulatinMU 0.097  0.068  0.080  0.092                             
## PopulatinRE 0.081  0.057  0.068  0.077  0.078                      
## PopulatinUM 0.083  0.058  0.069  0.078  0.079  0.067               
## PopulatinVA 0.102  0.072  0.085  0.097  0.098  0.083  0.084        
## PopulatinYE 0.085  0.060  0.070  0.080  0.081  0.068  0.069  0.086

TL_M_lmer_Kozeretska <- lmer(TL_micrometers ~ -1 + Population + (1|Line:Population) +
                               (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_TL_M,Supervisor.PI=='Kozeretska')))
capture.output(summary(TL_M_lmer_Kozeretska),file = "ThoraxLength/TL_M_Kozeretska_sum.txt")
capture.output(anova(TL_M_lmer_Kozeretska),file = "ThoraxLength/TL_M_lmer_Kozeretska.txt")
capture.output(emmeans(TL_M_lmer_Kozeretska, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ThoraxLength/TL_M_tk.txt")

anova(TL_M_lmer_Kozeretska)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 86499621 9611069     9 31.607  5283.6 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(TL_M_lmer_Kozeretska)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TL_micrometers ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_TL_M, Supervisor.PI == "Kozeretska"))
## 
## REML criterion at convergence: 50338.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.8281 -0.5822  0.0434  0.6572  7.9545 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  902.20  30.037  
##  Line:Population               (Intercept)  672.50  25.933  
##  Batch                         (Intercept)   11.11   3.332  
##  Residual                                  1819.04  42.650  
## Number of obs: 4807, groups:  
## ReplicateVial:Line:Population, 241; Line:Population, 130; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  835.238      8.442  38.650   98.94   <2e-16 ***
## PopulationGI  831.229     12.031  83.536   69.09   <2e-16 ***
## PopulationKA  812.504     10.148  59.023   80.07   <2e-16 ***
## PopulationMA  862.800      8.922  45.162   96.70   <2e-16 ***
## PopulationMU  838.174      8.766  42.553   95.62   <2e-16 ***
## PopulationRE  859.844     10.396  64.279   82.71   <2e-16 ***
## PopulationUM  838.730     10.270  61.509   81.67   <2e-16 ***
## PopulationVA  858.463      8.254  35.756  104.01   <2e-16 ***
## PopulationYE  845.745     10.109  61.155   83.66   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.055                                                  
## PopulatinKA 0.065  0.045                                           
## PopulatinMA 0.075  0.052  0.061                                    
## PopulatinMU 0.075  0.053  0.062  0.071                             
## PopulatinRE 0.063  0.044  0.053  0.060  0.061                      
## PopulatinUM 0.064  0.045  0.053  0.061  0.062  0.052               
## PopulatinVA 0.080  0.056  0.066  0.076  0.077  0.065  0.066        
## PopulatinYE 0.066  0.046  0.054  0.062  0.063  0.053  0.054  0.067

p_TL_F_Kozeretska <- ggplot(data = (subset(d_TL_F,Supervisor.PI=='Kozeretska')), aes(x=Population_Lat, y=TL_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_F_Kozeretska", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

p_TL_F_pop_Kozeretska <- ggplot(data = (subset(d_TL_F,Supervisor.PI=='Kozeretska')), aes(x=Population_Lat, y=TL_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_F_pop_Kozeretska", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

p_TL_M_Kozeretska <- ggplot(data = (subset(d_TL_M,Supervisor.PI=='Kozeretska')), aes(x=Population_Lat, y=TL_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_M_Kozeretska", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

p_TL_M_pop_Kozeretska <- ggplot(data = (subset(d_TL_M,Supervisor.PI=='Kozeretska')), aes(x=Population_Lat, y=TL_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_M_pop_Kozeretska", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

Posnien Lab

TL_F_lmer_Posnien <- lmer(TL_micrometers ~ -1 + Population + (1|Line:Population)+ 
                            (1|ReplicateVial : Line : Population), data = (subset(d_TL_F,Supervisor.PI=='Posnien')))
capture.output(summary(TL_F_lmer_Posnien),file = "ThoraxLength/TL_F_Posnien_sum.txt")
capture.output(anova(TL_F_lmer_Posnien),file = "ThoraxLength/TL_F_lmer_Posnien.txt")
capture.output(emmeans(TL_F_lmer_Posnien, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ThoraxLength/TL_F_lmer_Posnientk.txt")

anova(TL_F_lmer_Posnien)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 29385611 3265068     9    18  2123.9 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(TL_F_lmer_Posnien)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TL_micrometers ~ -1 + Population + (1 | Line:Population) + (1 |  
##     ReplicateVial:Line:Population)
##    Data: (subset(d_TL_F, Supervisor.PI == "Posnien"))
## 
## REML criterion at convergence: 2722.8
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.86894 -0.60033  0.05782  0.63035  2.78455 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  Line:Population               (Intercept)  542.6   23.29   
##  ReplicateVial:Line:Population (Intercept)  463.9   21.54   
##  Residual                                  1537.3   39.21   
## Number of obs: 270, groups:  
## Line:Population, 27; ReplicateVial:Line:Population, 27
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   880.66      19.67  18.00   44.78   <2e-16 ***
## PopulationGI   944.38      19.67  18.00   48.02   <2e-16 ***
## PopulationKA   928.61      19.67  18.00   47.22   <2e-16 ***
## PopulationMA   882.00      19.67  18.00   44.85   <2e-16 ***
## PopulationMU   866.23      19.67  18.00   44.05   <2e-16 ***
## PopulationRE   902.77      19.67  18.00   45.91   <2e-16 ***
## PopulationUM   908.43      19.67  18.00   46.19   <2e-16 ***
## PopulationVA   898.43      19.67  18.00   45.69   <2e-16 ***
## PopulationYE   941.85      19.67  18.00   47.89   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

TL_M_lmer_Posnien <- lmer(TL_micrometers ~ -1 + Population + (1|Line:Population), data = (subset(d_TL_M,Supervisor.PI=='Posnien')))
capture.output(summary(TL_M_lmer_Posnien),file = "ThoraxLength/TL_M_Posnien_sum.txt")
capture.output(anova(TL_M_lmer_Posnien),file = "ThoraxLength/TL_M_lmer_Posnien.txt")
capture.output(emmeans(TL_M_lmer_Posnien, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ThoraxLength/TL_M_lmer_Posnien_tk.txt")

anova(TL_M_lmer_Posnien)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 22599074 2511008     9    18  2066.5 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(TL_M_lmer_Posnien)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TL_micrometers ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_TL_M, Supervisor.PI == "Posnien"))
## 
## REML criterion at convergence: 2661.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.7213 -0.5390 -0.0356  0.5975  3.7202 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  814.9   28.55   
##  Residual                    1215.1   34.86   
## Number of obs: 270, groups:  Line:Population, 27
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   777.16      17.67  18.00   43.99   <2e-16 ***
## PopulationGI   841.23      17.67  18.00   47.61   <2e-16 ***
## PopulationKA   823.56      17.67  18.00   46.61   <2e-16 ***
## PopulationMA   785.43      17.67  18.00   44.46   <2e-16 ***
## PopulationMU   756.66      17.67  18.00   42.83   <2e-16 ***
## PopulationRE   802.59      17.67  18.00   45.43   <2e-16 ***
## PopulationUM   821.89      17.67  18.00   46.52   <2e-16 ***
## PopulationVA   796.67      17.67  18.00   45.09   <2e-16 ***
## PopulationYE   819.56      17.67  18.00   46.39   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_TL_F_Posnien <- ggplot(data = (subset(d_TL_F,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=TL_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_F_Posnien", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

p_TL_F_pop_Posnien <- ggplot(data = (subset(d_TL_F,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=TL_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_F_pop_Posnien", x="Population", y = "TL_micrometers")+ ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

p_TL_M_Posnien <- ggplot(data = (subset(d_TL_M,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=TL_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_M_Posnien", x="Population", y = "TL_micrometers")+ ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

p_TL_M_pop_Posnien <- ggplot(data = (subset(d_TL_M,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=TL_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_M_pop_Posnien", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

Ritchie Lab

TL_F_lmer_Ritchie <- lmer(TL_micrometers ~ -1 + Population + (1|Line:Population) +
                            (1|Batch), data = (subset(d_TL_F,Supervisor.PI=='Ritchie')))
capture.output(summary(TL_F_lmer_Ritchie),file = "ThoraxLength/TL_F_Ritchie_sum.txt")
capture.output(anova(TL_F_lmer_Ritchie),file = "ThoraxLength/TL_F_lmer_Ritchie.txt")
capture.output(emmeans(TL_F_lmer_Ritchie, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ThoraxLength/TL_F_lmer_Ritchie_tk.txt")

anova(TL_F_lmer_Ritchie)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 24034212 4005702     6 12.87  1422.3 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(TL_F_lmer_Ritchie)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TL_micrometers ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: (subset(d_TL_F, Supervisor.PI == "Ritchie"))
## 
## REML criterion at convergence: 11444.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.7332 -0.5260  0.0394  0.6188  5.1914 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  776.8   27.87   
##  Batch           (Intercept)  913.4   30.22   
##  Residual                    2816.3   53.07   
## Number of obs: 1059, groups:  Line:Population, 26; Batch, 12
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   955.31      19.71  19.55   48.46   <2e-16 ***
## PopulationGI   969.43      18.08  18.38   53.62   <2e-16 ***
## PopulationKA   956.35      18.44  17.17   51.86   <2e-16 ***
## PopulationMA   976.97      18.40  17.01   53.11   <2e-16 ***
## PopulationMU   970.78      25.57  17.64   37.97   <2e-16 ***
## PopulationYE   988.71      17.62  19.32   56.10   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU
## PopulatinGI 0.446                             
## PopulatinKA 0.011  0.013                      
## PopulatinMA 0.011  0.013  0.501               
## PopulatinMU 0.053  0.063  0.202  0.203        
## PopulatinYE 0.300  0.353  0.036  0.036  0.177

TL_M_lmer_Ritchie <- lmer(TL_micrometers ~ -1 + Population + (1|Line:Population) +
                            (1|Batch), data = (subset(d_TL_M,Supervisor.PI=='Ritchie')))
capture.output(summary(TL_M_lmer_Ritchie),file = "ThoraxLength/TL_M_Ritchie_sum.txt")
capture.output(anova(TL_M_lmer_Ritchie),file = "ThoraxLength/TL_M_lmer_Ritchie.txt")
capture.output(emmeans(TL_M_lmer_Ritchie, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ThoraxLength/TL_M_lmer_Ritchie_tk.txt")

anova(TL_M_lmer_Ritchie)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 16132204 2688701     6 12.765  751.17 1.654e-15 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(TL_M_lmer_Ritchie)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TL_micrometers ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: (subset(d_TL_M, Supervisor.PI == "Ritchie"))
## 
## REML criterion at convergence: 11309.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -8.8016 -0.5065  0.0525  0.5479  3.8390 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  824.6   28.72   
##  Batch           (Intercept) 1396.0   37.36   
##  Residual                    3579.3   59.83   
## Number of obs: 1024, groups:  Line:Population, 26; Batch, 12
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   826.89      22.01  19.34   37.56   <2e-16 ***
## PopulationGI   823.04      20.37  18.13   40.40   <2e-16 ***
## PopulationKA   815.06      20.90  16.85   38.99   <2e-16 ***
## PopulationMA   840.52      20.91  16.88   40.20   <2e-16 ***
## PopulationMU   844.55      27.74  17.96   30.44   <2e-16 ***
## PopulationYE   846.22      19.75  19.14   42.84   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU
## PopulatinGI 0.523                             
## PopulatinKA 0.020  0.023                      
## PopulatinMA 0.020  0.023  0.581               
## PopulatinMU 0.080  0.092  0.252  0.252        
## PopulatinYE 0.372  0.427  0.054  0.054  0.215

p_TL_F_Ritchie <- filter(d_TL_F, Supervisor.PI == "Ritchie") %>% 
  ggplot(aes(x=Population_Lat, y=TL_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_F_Ritchie", x="Population", y = "TL_micrometers") +
  ylim(c(min_TL, max_TL)) + 
  droseu_theme

## quartz_off_screen 
##                 2

p_TL_F_pop_Ritchie <- ggplot(data = (subset(d_TL_F,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=TL_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_F_pop_Ritchie", x="Population", y = "TL_micrometers")+ coord_cartesian(ylim =c(min_TL, max_TL))+ droseu_theme
dev.off()

## null device 
##           1

## quartz_off_screen 
##                 2

p_TL_M_Ritchie <- ggplot(data = (subset(d_TL_M,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=TL_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_M_Ritchie", x="Population", y = "TL_micrometers")+ ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

p_TL_M_pop_Ritchie <- ggplot(data = (subset(d_TL_M,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=TL_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_M_pop_Ritchie", x="Population", y = "TL_micrometers")+ ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

Schmidt Lab

TL_F_lmer_Schmidt <- lmer(TL_micrometers ~ -1 + Population + (1|Line:Population), data = (subset(d_TL_F,Supervisor.PI=='Schmidt')))
capture.output(summary(TL_F_lmer_Schmidt),file = "ThoraxLength/TL_F_Schmidt_sum.txt")
capture.output(anova(TL_F_lmer_Schmidt),file = "ThoraxLength/TL_F_lmer_Schmidt.txt")
capture.output(emmeans(TL_F_lmer_Schmidt, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ThoraxLength/TL_F_lmer_Schmidt_tk.txt")

anova(TL_F_lmer_Schmidt)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 355502327 39500259     9 146.65   31323 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(TL_F_lmer_Schmidt)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TL_micrometers ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_TL_F, Supervisor.PI == "Schmidt"))
## 
## REML criterion at convergence: 15278.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -8.8431 -0.5418  0.0956  0.6004  2.7028 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  436.8   20.90   
##  Residual                    1261.1   35.51   
## Number of obs: 1514, groups:  Line:Population, 157
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  995.047      5.305  143.475   187.6   <2e-16 ***
## PopulationGI 1021.950      6.661  149.941   153.4   <2e-16 ***
## PopulationKA 1005.383      5.440  143.154   184.8   <2e-16 ***
## PopulationMA 1014.892      5.487  147.165   185.0   <2e-16 ***
## PopulationMU 1004.770      5.318  144.672   189.0   <2e-16 ***
## PopulationRE 1036.717      6.446  152.174   160.8   <2e-16 ***
## PopulationUM 1011.081      6.153  145.685   164.3   <2e-16 ***
## PopulationVA 1012.334      5.764  144.330   175.6   <2e-16 ***
## PopulationYE 1009.885      5.367  149.854   188.2   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_TL_F_Schmidt <- ggplot(data = (subset(d_TL_F,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=TL_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_F_Schmidt", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

p_TL_F_pop_Schmidt <- ggplot(data = (subset(d_TL_F,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=TL_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_TL_F_pop_Schmidt", x="Population", y = "TL_micrometers") + ylim(c(min_TL, max_TL))+ droseu_theme

## quartz_off_screen 
##                 2

Lab correlations

intersect_Line_TL  <- Reduce(intersect, list(d_TL$Line[d_TL$Supervisor.PI == "Kozeretska"],
                                        d_TL$Line[d_TL$Supervisor.PI == "Posnien"],
                                        d_TL$Line[d_TL$Supervisor.PI == "Ritchie"],
                                        d_TL$Line[d_TL$Supervisor.PI == "Schmidt"]))

write(intersect_Line_TL, file ="ThoraxLength/intersect_Line_TL.txt", sep = "\t")

intersect_Pop_TL  <- Reduce(intersect, list(d_TL$Population[d_TL$Supervisor.PI == "Kozeretska"],
                                        d_TL$Population[d_TL$Supervisor.PI == "Posnien"],
                                        d_TL$Population[d_TL$Supervisor.PI == "Ritchie"],
                                        d_TL$Population[d_TL$Supervisor.PI == "Schmidt"]))

write(intersect_Pop_TL, file ="ThoraxLength/intersect_Pop_TL.txt", sep = "\t")

There are 4 populations and 6 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

TL_fitted <- combineFitted(labs = rep(c("Kozeretska", "Posnien", "Ritchie"), 2), sex = rep(c("F", "M"), each = 3), models = list(TL_F_lmer_Kozeretska, TL_F_lmer_Posnien, TL_F_lmer_Ritchie, TL_M_lmer_Kozeretska, TL_M_lmer_Posnien, TL_M_lmer_Ritchie)) %>%
  bind_rows(combineFitted(labs = "Schmidt", sex = "F", models = list(TL_F_lmer_Schmidt)))

p_TL_F_lab_cor <- function() scatterPlotMatrix(TL_fitted, sex = "F")
p_TL_M_lab_cor <- function() scatterPlotMatrix(TL_fitted, sex = "M")

p_TL_F_lab_cor()

pdf(file="ThoraxLength/p_TL_F_lab_cor.pdf",width=8, height=8)
p_TL_F_lab_cor()
invisible(dev.off())

p_TL_M_lab_cor()

pdf(file="ThoraxLength/p_TL_M_lab_cor.pdf",width=8, height=8)
p_TL_M_lab_cor()
invisible(dev.off())

Wing area

dir.create(file.path("WingArea"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Onder Lab : Cansu Aksoy, Ekin Demir, Ezgi Cobanoglu, Banu Sebnem Onder

Posnien Lab : Micael Reis, Lennart Hüper, Nico Posnien

Ritchie Lab : Megan Mcgunnigle, Nicola Cook, Teresa Abaurrea, Marija Tanaskovic, Michael Ritchie

Stamenkovic-Radak Lab : Marija Savic Veselinovic, Marija Tanaskovic, Aleksandra Patenkovic, Filip Filopovski, Mihailo Jelic, Katarina Eric, Pavle Eric, Slobodan Davidovic, Marina Stamenkovic-Radak

Data Reformatting

d_WA <- read.csv("MasterSheets_Oct21_git/WA_MasterSheet_Nov21.csv")
str(d_WA)

## 'data.frame':    21201 obs. of  15 variables:
##  $ Supervisor.PI                : chr  "StamenkovicRadak" "StamenkovicRadak" "StamenkovicRadak" "StamenkovicRadak" ...
##  $ Diet                         : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch                        : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population                   : chr  "YE" "YE" "YE" "YE" ...
##  $ Line                         : chr  "YE13" "YE13" "YE13" "YE13" ...
##  $ Sex                          : chr  "F" "F" "F" "F" ...
##  $ ReplicateVialOld             : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial                : chr  "StamenkovicRadak_1_YE13_1" "StamenkovicRadak_1_YE13_1" "StamenkovicRadak_1_YE13_1" "StamenkovicRadak_1_YE13_1" ...
##  $ Individual                   : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ CentroidSizeLeft_micrometers : num  2869 2910 2901 3050 2906 ...
##  $ CentroidSizeRight_micrometers: num  2891 2889 2881 3039 2919 ...
##  $ Country                      : chr  "Turkey" "Turkey" "Turkey" "Turkey" ...
##  $ Latitude                     : num  40.2 40.2 40.2 40.2 40.2 ...
##  $ Longitude                    : num  32.3 32.3 32.3 32.3 32.3 ...
##  $ Altitude                     : int  680 680 680 680 680 680 680 680 680 680 ...

d_WA$Supervisor.PI <- as.factor(d_WA$Supervisor.PI)
d_WA$Diet <- as.factor(d_WA$Diet)
d_WA$Batch <- as.factor(d_WA$Batch)
d_WA$Population_Lat <- factor(d_WA$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_WA$Population_Lon <- factor(d_WA$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_WA$Population_Alt <- factor(d_WA$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_WA$Line <- as.factor(d_WA$Line)
d_WA$Sex <- as.factor(d_WA$Sex)
d_WA$ReplicateVial <- as.factor(d_WA$ReplicateVial)
d_WA$CentroidSizeLeft_micrometers <- as.numeric(d_WA$CentroidSizeLeft_micrometers)
d_WA$CentroidSizeRight_micrometers <- as.numeric(d_WA$CentroidSizeRight_micrometers)

str(d_WA)

## 'data.frame':    21201 obs. of  18 variables:
##  $ Supervisor.PI                : Factor w/ 4 levels "Onder","Posnien",..: 4 4 4 4 4 4 4 4 4 4 ...
##  $ Diet                         : Factor w/ 2 levels "NS","S": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch                        : Factor w/ 15 levels "1","2","3","4",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population                   : chr  "YE" "YE" "YE" "YE" ...
##  $ Line                         : Factor w/ 172 levels "AK1","AK10","AK11",..: 154 154 154 154 154 154 154 154 154 154 ...
##  $ Sex                          : Factor w/ 2 levels "F","M": 1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVialOld             : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial                : Factor w/ 806 levels "Onder_1_AK1_1",..: 711 711 711 711 711 711 711 711 711 711 ...
##  $ Individual                   : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ CentroidSizeLeft_micrometers : num  2869 2910 2901 3050 2906 ...
##  $ CentroidSizeRight_micrometers: num  2891 2889 2881 3039 2919 ...
##  $ Country                      : chr  "Turkey" "Turkey" "Turkey" "Turkey" ...
##  $ Latitude                     : num  40.2 40.2 40.2 40.2 40.2 ...
##  $ Longitude                    : num  32.3 32.3 32.3 32.3 32.3 ...
##  $ Altitude                     : int  680 680 680 680 680 680 680 680 680 680 ...
##  $ Population_Lat               : Factor w/ 9 levels "YE","RE","GI",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population_Lon               : Factor w/ 9 levels "RE","GI","KA",..: 8 8 8 8 8 8 8 8 8 8 ...
##  $ Population_Alt               : Factor w/ 9 levels "KA","AK","GI",..: 9 9 9 9 9 9 9 9 9 9 ...

d_WA_F <-subset(d_WA,Sex=='F')
d_WA_M <-subset(d_WA,Sex=='M')

Left

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

Descriptive statistics at the line level, without batch information :

Descriptive statistics at the population level, with batch information :

Descriptive statistics at the population level, without batch information :

Plots and Linear Models by Lab

Overview

min_WAL <- min(d_WA$CentroidSizeLeft_micrometers)
max_WAL <- max(d_WA$CentroidSizeLeft_micrometers)

y-axis is scaled by the minimum (1857.612491) and maximum (3333.0502) values in the full data set.

p_WA_F_L <- ggplot(data = d_WA_F, aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Supervisor.PI)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_L", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L <- ggplot(data = d_WA_M, aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Supervisor.PI)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_L", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

Onder Lab

WA_F_L_Onder_lmer <- lmer(CentroidSizeLeft_micrometers ~ -1 + Population + (1|Line:Population) + (1|Batch) + (1|ReplicateVial:Line:Population), data = d_WA_F %>% filter(Supervisor.PI=='Onder'))
capture.output(summary(WA_F_L_Onder_lmer),file = "WingArea/WA_F_L_Onder_sum.txt")
capture.output(anova(WA_F_L_Onder_lmer),file = "WingArea/WA_F_L_Onder_lmer.txt")
capture.output(emmeans(WA_F_L_Onder_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_L_Onder_lmer_tk.txt")

anova(WA_F_L_Onder_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 233520534 25946726     9 25.192  4460.1 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_F_L_Onder_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeLeft_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: d_WA_F %>% filter(Supervisor.PI == "Onder")
## 
## REML criterion at convergence: 61105.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.7289 -0.5621  0.0553  0.6206  4.0410 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  781.8   27.96   
##  Line:Population               (Intercept) 3091.0   55.60   
##  Batch                         (Intercept) 1209.8   34.78   
##  Residual                                  5817.5   76.27   
## Number of obs: 5247, groups:  
## ReplicateVial:Line:Population, 518; Line:Population, 167; Batch, 7
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2813.82      19.10   14.95   147.3   <2e-16 ***
## PopulationGI  2852.89      20.67   19.89   138.0   <2e-16 ***
## PopulationKA  2858.67      19.15   15.07   149.2   <2e-16 ***
## PopulationMA  2898.81      19.20   15.14   151.0   <2e-16 ***
## PopulationMU  2860.10      18.91   14.53   151.3   <2e-16 ***
## PopulationRE  2943.87      20.22   18.43   145.6   <2e-16 ***
## PopulationUM  2901.10      19.80   17.13   146.5   <2e-16 ***
## PopulationVA  2843.50      19.07   14.84   149.1   <2e-16 ***
## PopulationYE  2815.30      19.23   15.38   146.4   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.467                                                  
## PopulatinKA 0.487  0.466                                           
## PopulatinMA 0.516  0.469  0.488                                    
## PopulatinMU 0.504  0.462  0.497  0.505                             
## PopulatinRE 0.472  0.444  0.478  0.472  0.475                      
## PopulatinUM 0.474  0.447  0.487  0.473  0.478  0.460               
## PopulatinVA 0.505  0.470  0.507  0.508  0.507  0.479  0.483        
## PopulatinYE 0.500  0.463  0.497  0.500  0.499  0.473  0.477  0.503

WA_M_L_Onder_lmer <- lmer(CentroidSizeLeft_micrometers ~ -1 + Population + (1|Line:Population) +
                            (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_M,Supervisor.PI=='Onder')))
capture.output(summary(WA_M_L_Onder_lmer),file = "WingArea/WA_M_L_Onder_sum.txt")
capture.output(anova(WA_M_L_Onder_lmer),file = "WingArea/WA_M_L_Onder_lmer.txt")
capture.output(emmeans(WA_M_L_Onder_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_M_L_Onder_lmer_tk.txt")

anova(WA_M_L_Onder_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 418745747 46527305     9 25.189    9902 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_M_L_Onder_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeLeft_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_M, Supervisor.PI == "Onder"))
## 
## REML criterion at convergence: 59806.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.2993 -0.5880  0.0332  0.6103  5.8833 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  765.5   27.67   
##  Line:Population               (Intercept) 2085.2   45.66   
##  Batch                         (Intercept)  336.6   18.35   
##  Residual                                  4698.8   68.55   
## Number of obs: 5230, groups:  
## ReplicateVial:Line:Population, 518; Line:Population, 167; Batch, 7
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2444.13      13.49   29.28   181.2   <2e-16 ***
## PopulationGI  2476.23      15.02   41.35   164.9   <2e-16 ***
## PopulationKA  2478.28      13.53   29.32   183.1   <2e-16 ***
## PopulationMA  2509.23      13.57   29.21   184.9   <2e-16 ***
## PopulationMU  2479.17      13.31   29.20   186.3   <2e-16 ***
## PopulationRE  2545.32      14.62   38.49   174.1   <2e-16 ***
## PopulationUM  2513.23      14.19   35.53   177.1   <2e-16 ***
## PopulationVA  2488.63      13.45   28.93   185.0   <2e-16 ***
## PopulationYE  2440.14      13.63   30.92   179.0   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.271                                                  
## PopulatinKA 0.279  0.271                                           
## PopulatinMA 0.317  0.275  0.282                                    
## PopulatinMU 0.298  0.263  0.290  0.301                             
## PopulatinRE 0.272  0.254  0.280  0.273  0.274                      
## PopulatinUM 0.271  0.254  0.288  0.270  0.273  0.265               
## PopulatinVA 0.301  0.274  0.305  0.306  0.301  0.280  0.281        
## PopulatinYE 0.296  0.268  0.293  0.298  0.293  0.275  0.276  0.300

p_WA_F_L_Onder <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_Onder_L", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_F_L_pop_Onder <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_L_pop_Onder", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L_Onder <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_Onder_L", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L_pop_Onder <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_L_pop_Onder", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

Posnien Lab

WA_F_L_Posnien_lmer <- lmer(CentroidSizeLeft_micrometers ~ -1 + Population + (1|Line:Population) +
                            (1|ReplicateVial : Line : Population), data = (subset(d_WA_F,Supervisor.PI=='Posnien')))

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## unable to evaluate scaled gradient

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge: degenerate Hessian with 1 negative eigenvalues

capture.output(summary(WA_F_L_Posnien_lmer),file = "WingArea/WA_F_L_Posnien_sum.txt")
capture.output(anova(WA_F_L_Posnien_lmer),file = "WingArea/WA_F_L_Posnien_lmer.txt")
capture.output(emmeans(WA_F_L_Posnien_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_L_Posnien_lmer_tk.txt")

anova(WA_F_L_Posnien_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 184927442 20547494     9    18  2460.7 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_F_L_Posnien_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeLeft_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_F, Supervisor.PI == "Posnien"))
## 
## REML criterion at convergence: 3169.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.3762 -0.5835 -0.0034  0.6637  2.5365 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  Line:Population               (Intercept)  432.2   20.79   
##  ReplicateVial:Line:Population (Intercept) 7243.4   85.11   
##  Residual                                  8350.2   91.38   
## Number of obs: 270, groups:  
## Line:Population, 27; ReplicateVial:Line:Population, 27
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2589.01      53.26   18.00   48.61   <2e-16 ***
## PopulationGI  2728.20      53.26   18.00   51.22   <2e-16 ***
## PopulationKA  2639.38      53.26   18.00   49.55   <2e-16 ***
## PopulationMA  2658.50      53.26   18.00   49.91   <2e-16 ***
## PopulationMU  2563.45      53.26   18.00   48.13   <2e-16 ***
## PopulationRE  2723.93      53.26   18.00   51.14   <2e-16 ***
## PopulationUM  2663.03      53.26   18.00   50.00   <2e-16 ***
## PopulationVA  2549.58      53.26   18.00   47.87   <2e-16 ***
## PopulationYE  2657.72      53.26   18.00   49.90   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## unable to evaluate scaled gradient
## Model failed to converge: degenerate  Hessian with 1 negative eigenvalues

WA_M_L_Posnien_lmer <- lmer(CentroidSizeLeft_micrometers ~ -1 + Population + (1|Line:Population) +
                              (1|ReplicateVial : Line : Population), data = (subset(d_WA_M,Supervisor.PI=='Posnien')))

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## unable to evaluate scaled gradient

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge: degenerate Hessian with 1 negative eigenvalues

## Warning: Model failed to converge with 1 negative eigenvalue: -2.6e-05

capture.output(summary(WA_M_L_Posnien_lmer),file = "WingArea/WA_M_L_Posnien_sum.txt")
capture.output(anova(WA_M_L_Posnien_lmer),file = "WingArea/WA_M_L_Posnien_lmer.txt")
capture.output(emmeans(WA_M_L_Posnien_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_L_Posnien_lmer_tk.txt")

anova(WA_M_L_Posnien_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 120639111 13404346     9    18  2235.9 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_M_L_Posnien_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeLeft_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_M, Supervisor.PI == "Posnien"))
## 
## REML criterion at convergence: 3086.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.9072 -0.4977  0.0192  0.6488  3.4055 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  Line:Population               (Intercept)   65.11   8.069  
##  ReplicateVial:Line:Population (Intercept) 6715.55  81.948  
##  Residual                                  5995.06  77.428  
## Number of obs: 270, groups:  
## Line:Population, 27; ReplicateVial:Line:Population, 27
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   2300.0       49.6   18.0   46.37   <2e-16 ***
## PopulationGI   2411.8       49.6   18.0   48.63   <2e-16 ***
## PopulationKA   2332.6       49.6   18.0   47.03   <2e-16 ***
## PopulationMA   2381.3       49.6   18.0   48.01   <2e-16 ***
## PopulationMU   2266.6       49.6   18.0   45.70   <2e-16 ***
## PopulationRE   2374.9       49.6   18.0   47.88   <2e-16 ***
## PopulationUM   2348.3       49.6   18.0   47.35   <2e-16 ***
## PopulationVA   2339.7       49.6   18.0   47.17   <2e-16 ***
## PopulationYE   2349.2       49.6   18.0   47.36   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## unable to evaluate scaled gradient
## Model failed to converge: degenerate  Hessian with 1 negative eigenvalues

p_WA_F_L_Posnien <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_L_Posnien", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_F_L_pop_Posnien <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_L_pop_Posnien", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L_Posnien <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_L_Posnien", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L_pop_Posnien <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_L_pop_Posnien", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

Ritchie Lab

WA_F_L_Ritchie_lmer <- lmer(CentroidSizeLeft_micrometers ~ -1 + Population + (1|Line:Population) +
                              (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_F,Supervisor.PI=='Ritchie')))
capture.output(summary(WA_F_L_Ritchie_lmer),file = "WingArea/WA_F_L_Ritchie_sum.txt")
capture.output(anova(WA_F_L_Ritchie_lmer),file = "WingArea/WA_F_L_Ritchie_lmer.txt")
capture.output(emmeans(WA_F_L_Ritchie_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_L_Ritchie_lmer_tk.txt")

anova(WA_F_L_Ritchie_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 120238982 13359887     9 24.415    1522 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_F_L_Ritchie_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeLeft_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_F, Supervisor.PI == "Ritchie"))
## 
## REML criterion at convergence: 16563.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.3958 -0.5231  0.0337  0.6233  3.1535 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  6184     78.64  
##  Line:Population               (Intercept) 17974    134.07  
##  Batch                         (Intercept)  1663     40.79  
##  Residual                                   8778     93.69  
## Number of obs: 1367, groups:  
## ReplicateVial:Line:Population, 132; Line:Population, 47; Batch, 15
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2759.24      74.08   27.97   37.24   <2e-16 ***
## PopulationGI  2587.18      56.41   22.96   45.86   <2e-16 ***
## PopulationKA  2714.14      73.97   27.72   36.69   <2e-16 ***
## PopulationMA  2846.67      66.22   27.51   42.99   <2e-16 ***
## PopulationMU  2782.90      66.15   27.67   42.07   <2e-16 ***
## PopulationRE  2878.60      66.08   27.75   43.56   <2e-16 ***
## PopulationUM  2889.01      66.15   27.71   43.67   <2e-16 ***
## PopulationVA  2774.25      66.16   27.77   41.93   <2e-16 ***
## PopulationYE  2782.82      66.14   27.83   42.08   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.056                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.066                                    
## PopulatinMU 0.001  0.001  0.050  0.060                             
## PopulatinRE 0.001  0.001  0.033  0.030  0.017                      
## PopulatinUM 0.002  0.003  0.002  0.002  0.015  0.045               
## PopulatinVA 0.017  0.013  0.002  0.001  0.002  0.045  0.059        
## PopulatinYE 0.035  0.064  0.001  0.001  0.015  0.002  0.029  0.017

WA_M_L_Ritchie_lmer <- lmer(CentroidSizeLeft_micrometers ~ -1 + Population + (1|Line:Population) +
                              (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_M,Supervisor.PI=='Ritchie')))
capture.output(summary(WA_M_L_Ritchie_lmer),file = "WingArea/WA_M_L_Ritchie_sum.txt")
capture.output(anova(WA_M_L_Ritchie_lmer),file = "WingArea/WA_M_L_Ritchie_lmer.txt")
capture.output(emmeans(WA_M_L_Ritchie_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_M_L_Ritchie_lmer_tk.txt")

anova(WA_M_L_Ritchie_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 177892587 19765843     9 22.802  2847.6 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_M_L_Ritchie_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeLeft_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_M, Supervisor.PI == "Ritchie"))
## 
## REML criterion at convergence: 16229.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.9391 -0.4829  0.0696  0.5871  4.6442 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept) 4539.9   67.38   
##  Line:Population               (Intercept) 6325.9   79.54   
##  Batch                         (Intercept)  740.4   27.21   
##  Residual                                  6941.1   83.31   
## Number of obs: 1368, groups:  
## ReplicateVial:Line:Population, 132; Line:Population, 44; Batch, 15
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2448.31      46.86   35.17   52.25   <2e-16 ***
## PopulationGI  2383.26      41.92   34.95   56.85   <2e-16 ***
## PopulationKA  2425.85      41.94   35.01   57.84   <2e-16 ***
## PopulationMA  2478.01      41.91   34.92   59.12   <2e-16 ***
## PopulationMU  2436.41      41.85   35.01   58.21   <2e-16 ***
## PopulationRE  2495.99      41.75   34.81   59.78   <2e-16 ***
## PopulationUM  2511.63      41.87   35.08   59.99   <2e-16 ***
## PopulationVA  2432.26      41.88   35.14   58.08   <2e-16 ***
## PopulationYE  2413.16      41.87   35.16   57.64   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.073                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.081                                    
## PopulatinMU 0.001  0.001  0.067  0.067                             
## PopulatinRE 0.001  0.001  0.034  0.034  0.019                      
## PopulatinUM 0.002  0.003  0.003  0.003  0.017  0.050               
## PopulatinVA 0.019  0.018  0.002  0.002  0.002  0.050  0.065        
## PopulatinYE 0.039  0.051  0.001  0.001  0.016  0.003  0.033  0.019

p_WA_F_L_Ritchie <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_L_Ritchie", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_F_L_pop_Ritchie <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_L_pop_Ritchie", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L_Ritchie <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_L_Ritchie", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L_pop_Ritchie <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_L_pop_Ritchie", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

Stamenkovic-Radak Lab

WA_F_L_StamenkovicRadak_lmer <- lmer(CentroidSizeLeft_micrometers ~ -1 + Population + (1|Line:Population) +
                                       (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_F,Supervisor.PI=='StamenkovicRadak')))

## boundary (singular) fit: see ?isSingular

## Warning: Model failed to converge with 1 negative eigenvalue: -7.2e-03

capture.output(summary(WA_F_L_StamenkovicRadak_lmer),file = "WingArea/WA_F_L_StamenkovicRadak_sum.txt")
capture.output(anova(WA_F_L_StamenkovicRadak_lmer),file = "WingArea/WA_F_L_StamenkovicRadak_lmer.txt")
capture.output(emmeans(WA_F_L_StamenkovicRadak_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_L_StamenkovicRadak_lmer_tk.txt")

anova(WA_F_L_StamenkovicRadak_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##                Sum Sq   Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2879060673 319895630     9 114.59   41324 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_F_L_StamenkovicRadak_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeLeft_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_F, Supervisor.PI == "StamenkovicRadak"))
## 
## REML criterion at convergence: 44876.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.4785 -0.5666  0.0599  0.6399  3.6361 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)    6.451  2.54   
##  Line:Population               (Intercept) 2434.129 49.34   
##  Batch                         (Intercept)    0.000  0.00   
##  Residual                                  7741.151 87.98   
## Number of obs: 3787, groups:  
## ReplicateVial:Line:Population, 124; Line:Population, 124; Batch, 5
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2780.82      12.95  113.51   214.7   <2e-16 ***
## PopulationGI  2836.04      16.43  114.87   172.6   <2e-16 ***
## PopulationKA  2810.31      13.01  115.62   216.0   <2e-16 ***
## PopulationMA  2862.63      15.67  115.06   182.6   <2e-16 ***
## PopulationMU  2847.78      13.35  112.79   213.2   <2e-16 ***
## PopulationRE  2932.82      14.41  114.82   203.5   <2e-16 ***
## PopulationUM  2882.97      13.92  115.85   207.1   <2e-16 ***
## PopulationVA  2838.45      13.40  114.48   211.7   <2e-16 ***
## PopulationYE  2822.12      13.87  114.42   203.4   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

WA_M_L_StamenkovicRadak_lmer <- lmer(CentroidSizeLeft_micrometers ~ -1 + Population + (1|Line:Population) +
                                       (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_M,Supervisor.PI=='StamenkovicRadak')))

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## unable to evaluate scaled gradient

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge: degenerate Hessian with 1 negative eigenvalues

## Warning: Model failed to converge with 1 negative eigenvalue: -1.2e-03

capture.output(summary(WA_M_L_StamenkovicRadak_lmer),file = "WingArea/WA_M_L_StamenkovicRadak_sum.txt")
capture.output(anova(WA_M_L_StamenkovicRadak_lmer),file = "WingArea/WA_M_L_StamenkovicRadak_lmer.txt")
capture.output(emmeans(WA_M_L_StamenkovicRadak_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_M_L_StamenkovicRadak_lmer_tk.txt")

anova(WA_M_L_StamenkovicRadak_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq   Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 1.992e+09 221338867     9 60.345   36999 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_M_L_StamenkovicRadak_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeLeft_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_M, Supervisor.PI == "StamenkovicRadak"))
## 
## REML criterion at convergence: 42455.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -7.2393 -0.5570  0.0480  0.6236  6.0054 
## 
## Random effects:
##  Groups                        Name        Variance  Std.Dev.
##  ReplicateVial:Line:Population (Intercept)    0.3221  0.5675 
##  Line:Population               (Intercept) 1857.6982 43.1010 
##  Batch                         (Intercept)    6.6839  2.5853 
##  Residual                                  5982.2379 77.3449 
## Number of obs: 3662, groups:  
## ReplicateVial:Line:Population, 124; Line:Population, 124; Batch, 5
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2415.63      11.41   58.45   211.7   <2e-16 ***
## PopulationGI  2472.14      14.42   88.81   171.4   <2e-16 ***
## PopulationKA  2434.72      11.45   70.34   212.7   <2e-16 ***
## PopulationMA  2490.11      13.76   91.97   181.0   <2e-16 ***
## PopulationMU  2468.42      11.76   70.58   209.9   <2e-16 ***
## PopulationRE  2537.58      12.70   80.56   199.8   <2e-16 ***
## PopulationUM  2507.21      12.23   77.98   205.0   <2e-16 ***
## PopulationVA  2461.46      11.81   66.21   208.5   <2e-16 ***
## PopulationYE  2447.49      12.22   77.61   200.3   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.010                                                  
## PopulatinKA 0.013  0.010                                           
## PopulatinMA 0.009  0.008  0.010                                    
## PopulatinMU 0.013  0.009  0.012  0.009                             
## PopulatinRE 0.012  0.008  0.011  0.008  0.012                      
## PopulatinUM 0.011  0.009  0.012  0.009  0.012  0.011               
## PopulatinVA 0.014  0.010  0.012  0.009  0.013  0.011  0.011        
## PopulatinYE 0.012  0.009  0.012  0.009  0.012  0.011  0.011  0.012 
## optimizer (nloptwrap) convergence code: 0 (OK)
## unable to evaluate scaled gradient
## Model failed to converge: degenerate  Hessian with 1 negative eigenvalues

p_WA_F_L_StamenkovicRadak <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_L_StamenkovicRadak", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_F_L_pop_StamenkovicRadak <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_L_pop_StamenkovicRadak", x="Population", y = "CentroidSizeLeft_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L_StamenkovicRadak <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_L_StamenkovicRadak", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_L_pop_StamenkovicRadak <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=CentroidSizeLeft_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_L_pop_StamenkovicRadak", x="Population", y = "CentroidSizeLeft_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## quartz_off_screen 
##                 2

Right

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

Descriptive statistics at the line level, without batch information :

Descriptive statistics at the population level, with batch information :

Descriptive statistics at the population level, without batch information :

Plots and Linear Models by Lab

Overview

min_WAR <- min(d_WA$CentroidSizeRight_micrometers)
max_WAR <- max(d_WA$CentroidSizeRight_micrometers)

p_WA_F_R <- ggplot(data = d_WA_F, aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Supervisor.PI)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_R", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAL, max_WAL))+ droseu_theme

## Warning: Removed 2 rows containing non-finite values (stat_boxplot).

## Warning: Removed 2 rows containing non-finite values (stat_boxplot).

## quartz_off_screen 
##                 2

p_WA_M_R <- ggplot(data = d_WA_M, aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Supervisor.PI)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R", x="Population", y = "CentroidSizeRight_micrometers")+ ylim(c(min_WAL, max_WAL))+ droseu_theme

## Warning: Removed 1 rows containing non-finite values (stat_boxplot).

## Warning: Removed 1 rows containing non-finite values (stat_boxplot).

## quartz_off_screen 
##                 2

Onder Lab

for (i in dev.list()[1]:dev.list()[length(dev.list())]) {
   dev.off()
}
# crashed bc tries to save lots -- need to remove some 

WA_F_R_Onder_lmer <- lmer(CentroidSizeRight_micrometers ~ -1 + Population + (1|Line:Population) +
                            (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_F,Supervisor.PI=='Onder')))
capture.output(summary(WA_F_R_Onder_lmer),file = "WingArea/WA_F_R_Onder_sum.txt")
capture.output(anova(WA_F_R_Onder_lmer),file = "WingArea/WA_F_R_Onder_lmer.txt")
capture.output(emmeans(WA_F_R_Onder_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_R_Onder_lmer_tk.txt")

anova(WA_F_R_Onder_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 254276248 28252916     9 24.56  4858.7 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_F_R_Onder_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeRight_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_F, Supervisor.PI == "Onder"))
## 
## REML criterion at convergence: 61106
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.6211 -0.5609  0.0470  0.6255  3.7505 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  806.7   28.40   
##  Line:Population               (Intercept) 3014.8   54.91   
##  Batch                         (Intercept) 1097.1   33.12   
##  Residual                                  5814.9   76.26   
## Number of obs: 5247, groups:  
## ReplicateVial:Line:Population, 518; Line:Population, 167; Batch, 7
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2815.32      18.58   15.46   151.5   <2e-16 ***
## PopulationGI  2854.81      20.16   20.71   141.6   <2e-16 ***
## PopulationKA  2860.05      18.63   15.57   153.5   <2e-16 ***
## PopulationMA  2899.67      18.67   15.65   155.3   <2e-16 ***
## PopulationMU  2860.41      18.38   15.02   155.6   <2e-16 ***
## PopulationRE  2944.63      19.71   19.16   149.4   <2e-16 ***
## PopulationUM  2901.85      19.28   17.80   150.5   <2e-16 ***
## PopulationVA  2845.20      18.55   15.33   153.4   <2e-16 ***
## PopulationYE  2815.71      18.71   15.92   150.5   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.449                                                  
## PopulatinKA 0.468  0.448                                           
## PopulatinMA 0.498  0.451  0.470                                    
## PopulatinMU 0.485  0.443  0.478  0.487                             
## PopulatinRE 0.454  0.426  0.460  0.453  0.457                      
## PopulatinUM 0.455  0.429  0.469  0.454  0.459  0.442               
## PopulatinVA 0.487  0.452  0.489  0.490  0.488  0.461  0.465        
## PopulatinYE 0.481  0.445  0.478  0.482  0.480  0.455  0.459  0.485

WA_M_R_Onder_lmer <- lmer(CentroidSizeRight_micrometers ~ -1 + Population + (1|Line:Population) +
                            (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_M,Supervisor.PI=='Onder')))
capture.output(summary(WA_M_R_Onder_lmer),file = "WingArea/WA_M_R_Onder_sum.txt")
capture.output(anova(WA_M_R_Onder_lmer),file = "WingArea/WA_M_R_Onder_lmer.txt")
capture.output(emmeans(WA_M_R_Onder_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_M_R_Onder_lmer_tk.txt")

anova(WA_M_R_Onder_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 469778882 52197654     9 27.622   11136 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_M_R_Onder_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeRight_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_M, Supervisor.PI == "Onder"))
## 
## REML criterion at convergence: 59827.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.1368 -0.5848  0.0327  0.6114  6.0665 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  874.6   29.57   
##  Line:Population               (Intercept) 2102.5   45.85   
##  Batch                         (Intercept)  282.9   16.82   
##  Residual                                  4687.4   68.46   
## Number of obs: 5230, groups:  
## ReplicateVial:Line:Population, 518; Line:Population, 167; Batch, 7
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2444.99      13.29   35.83   184.0   <2e-16 ***
## PopulationGI  2477.84      14.87   50.37   166.6   <2e-16 ***
## PopulationKA  2481.34      13.34   35.75   186.1   <2e-16 ***
## PopulationMA  2509.80      13.37   35.50   187.7   <2e-16 ***
## PopulationMU  2479.30      13.11   36.10   189.2   <2e-16 ***
## PopulationRE  2547.09      14.46   47.19   176.1   <2e-16 ***
## PopulationUM  2513.79      14.02   43.85   179.3   <2e-16 ***
## PopulationVA  2489.41      13.25   35.41   187.9   <2e-16 ***
## PopulationYE  2441.24      13.44   38.10   181.7   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.238                                                  
## PopulatinKA 0.243  0.237                                           
## PopulatinMA 0.281  0.241  0.246                                    
## PopulatinMU 0.262  0.230  0.253  0.265                             
## PopulatinRE 0.238  0.222  0.246  0.238  0.240                      
## PopulatinUM 0.235  0.222  0.253  0.235  0.238  0.232               
## PopulatinVA 0.265  0.241  0.269  0.270  0.265  0.246  0.246        
## PopulatinYE 0.260  0.235  0.257  0.262  0.257  0.240  0.241  0.264

p_WA_F_R_Onder <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_Onder_R", x="Population", y = "CentroidSizeRight_micrometers")+ ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_F_R_pop_Onder <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_R_pop_Onder", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_R_Onder <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R_Onder", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_R_pop_Onder <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R_pop_Onder", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

Posnien Lab

WA_F_R_Posnien_lmer <- lmer(CentroidSizeRight_micrometers ~ -1 + Population + (1|Line:Population) +
                              (1|ReplicateVial : Line : Population), data = (subset(d_WA_F,Supervisor.PI=='Posnien')))

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## unable to evaluate scaled gradient

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge: degenerate Hessian with 1 negative eigenvalues

## Warning: Model failed to converge with 1 negative eigenvalue: -1.1e-04

capture.output(summary(WA_F_R_Posnien_lmer),file = "WingArea/WA_F_R_Posnien_sum.txt")
capture.output(anova(WA_F_R_Posnien_lmer),file = "WingArea/WA_F_R_Posnien_lmer.txt")
capture.output(emmeans(WA_F_R_Posnien_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_R_Posnien_lmer_tk.txt")

anova(WA_F_R_Posnien_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 189666658 21074073     9    18  2424.1 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_F_R_Posnien_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeRight_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_F, Supervisor.PI == "Posnien"))
## 
## REML criterion at convergence: 3180
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.88515 -0.60816  0.01116  0.64507  2.45427 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  Line:Population               (Intercept)  718.1   26.80   
##  ReplicateVial:Line:Population (Intercept) 7037.2   83.89   
##  Residual                                  8693.5   93.24   
## Number of obs: 270, groups:  
## Line:Population, 27; ReplicateVial:Line:Population, 27
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2584.98      53.62   18.00   48.21   <2e-16 ***
## PopulationGI  2725.24      53.62   18.00   50.83   <2e-16 ***
## PopulationKA  2637.33      53.62   18.00   49.19   <2e-16 ***
## PopulationMA  2657.28      53.62   18.00   49.56   <2e-16 ***
## PopulationMU  2562.32      53.62   18.00   47.79   <2e-16 ***
## PopulationRE  2720.04      53.62   18.00   50.73   <2e-16 ***
## PopulationUM  2660.50      53.62   18.00   49.62   <2e-16 ***
## PopulationVA  2547.93      53.62   18.00   47.52   <2e-16 ***
## PopulationYE  2657.20      53.62   18.00   49.56   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## unable to evaluate scaled gradient
## Model failed to converge: degenerate  Hessian with 1 negative eigenvalues

WA_M_R_Posnien_lmer <- lmer(CentroidSizeRight_micrometers ~ -1 + Population + (1|Line:Population) +
                               (1|ReplicateVial : Line : Population), data = (subset(d_WA_M,Supervisor.PI=='Posnien')))
capture.output(summary(WA_M_R_Posnien_lmer),file = "WingArea/WA_M_R_Posnien_sum.txt")
capture.output(anova(WA_M_R_Posnien_lmer),file = "WingArea/WA_M_R_Posnien_lmer.txt")
capture.output(emmeans(WA_M_R_Posnien_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_M_R_Posnien_lmer_tk.txt")

anova(WA_M_R_Posnien_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 118227620 13136402     9    18  2177.8 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_M_R_Posnien_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeRight_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_M, Supervisor.PI == "Posnien"))
## 
## REML criterion at convergence: 3088.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.0851 -0.5566 -0.0153  0.6569  3.5274 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  Line:Population               (Intercept)   38.41   6.198  
##  ReplicateVial:Line:Population (Intercept) 6927.30  83.230  
##  Residual                                  6032.05  77.666  
## Number of obs: 270, groups:  
## Line:Population, 27; ReplicateVial:Line:Population, 27
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2296.73      50.23   18.00   45.73   <2e-16 ***
## PopulationGI  2405.06      50.23   18.00   47.88   <2e-16 ***
## PopulationKA  2331.64      50.23   18.00   46.42   <2e-16 ***
## PopulationMA  2382.90      50.23   18.00   47.44   <2e-16 ***
## PopulationMU  2273.85      50.23   18.00   45.27   <2e-16 ***
## PopulationRE  2374.99      50.23   18.00   47.28   <2e-16 ***
## PopulationUM  2349.13      50.23   18.00   46.77   <2e-16 ***
## PopulationVA  2334.22      50.23   18.00   46.47   <2e-16 ***
## PopulationYE  2344.77      50.23   18.00   46.68   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_WA_F_R_Posnien <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_R_Posnien", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_F_R_pop_Posnien <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_R_pop_Posnien", x="Population", y = "CentroidSizeRight_micrometers")+ ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_R_Posnien <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R_Posnien", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_R_pop_Posnien <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Posnien')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R_pop_Posnien", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

Ritchie Lab

WA_F_R_Ritchie_lmer <- lmer(CentroidSizeRight_micrometers ~ -1 + Population + (1|Line:Population) +
                              (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_F,Supervisor.PI=='Ritchie')))
capture.output(summary(WA_F_R_Ritchie_lmer),file = "WingArea/WA_F_R_Ritchie_sum.txt")
capture.output(anova(WA_F_R_Ritchie_lmer),file = "WingArea/WA_F_R_Ritchie_lmer.txt")
capture.output(emmeans(WA_F_R_Ritchie_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_R_Ritchie_lmer_tk.txt")

anova(WA_F_R_Ritchie_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 124267590 13807510     9 24.263  1575.7 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_F_R_Ritchie_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeRight_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_F, Supervisor.PI == "Ritchie"))
## 
## REML criterion at convergence: 16564.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.4488 -0.5249  0.0615  0.6440  3.1927 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  6492     80.58  
##  Line:Population               (Intercept) 17468    132.17  
##  Batch                         (Intercept)  1632     40.39  
##  Residual                                   8763     93.61  
## Number of obs: 1367, groups:  
## ReplicateVial:Line:Population, 132; Line:Population, 47; Batch, 15
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2777.36      73.35   27.73   37.86   <2e-16 ***
## PopulationGI  2611.67      55.96   23.04   46.67   <2e-16 ***
## PopulationKA  2731.85      73.24   27.47   37.30   <2e-16 ***
## PopulationMA  2875.40      65.56   27.27   43.86   <2e-16 ***
## PopulationMU  2805.69      65.49   27.42   42.84   <2e-16 ***
## PopulationRE  2896.22      65.42   27.50   44.27   <2e-16 ***
## PopulationUM  2913.64      65.49   27.47   44.49   <2e-16 ***
## PopulationVA  2794.00      65.50   27.52   42.66   <2e-16 ***
## PopulationYE  2797.38      65.48   27.59   42.72   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.056                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.066                                    
## PopulatinMU 0.001  0.001  0.050  0.060                             
## PopulatinRE 0.001  0.001  0.033  0.030  0.017                      
## PopulatinUM 0.002  0.003  0.002  0.002  0.015  0.045               
## PopulatinVA 0.017  0.013  0.002  0.001  0.002  0.045  0.059        
## PopulatinYE 0.035  0.064  0.001  0.001  0.015  0.002  0.029  0.017

WA_M_R_Ritchie_lmer <- lmer(CentroidSizeRight_micrometers ~ -1 + Population + (1|Line:Population) +
                              (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_M,Supervisor.PI=='Ritchie')))
capture.output(summary(WA_M_R_Ritchie_lmer),file = "WingArea/WA_M_R_Ritchie_sum.txt")
capture.output(anova(WA_M_R_Ritchie_lmer),file = "WingArea/WA_M_R_Ritchie_lmer.txt")
capture.output(emmeans(WA_M_R_Ritchie_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_M_R_Ritchie_lmer_tk.txt")

anova(WA_M_R_Ritchie_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 185684975 20631664     9 22.81  3002.1 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_M_R_Ritchie_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeRight_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_M, Supervisor.PI == "Ritchie"))
## 
## REML criterion at convergence: 16215.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.9851 -0.5112  0.0484  0.5839  4.7200 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept) 4472.4   66.88   
##  Line:Population               (Intercept) 6265.6   79.16   
##  Batch                         (Intercept)  642.8   25.35   
##  Residual                                  6872.5   82.90   
## Number of obs: 1368, groups:  
## ReplicateVial:Line:Population, 132; Line:Population, 44; Batch, 15
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2462.73      46.40   35.18   53.08   <2e-16 ***
## PopulationGI  2401.15      41.50   34.98   57.87   <2e-16 ***
## PopulationKA  2445.67      41.51   35.04   58.91   <2e-16 ***
## PopulationMA  2499.15      41.49   34.95   60.24   <2e-16 ***
## PopulationMU  2457.23      41.44   35.02   59.30   <2e-16 ***
## PopulationRE  2516.54      41.35   34.82   60.87   <2e-16 ***
## PopulationUM  2532.84      41.45   35.09   61.10   <2e-16 ***
## PopulationVA  2448.82      41.47   35.15   59.06   <2e-16 ***
## PopulationYE  2431.57      41.46   35.17   58.65   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.065                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.072                                    
## PopulatinMU 0.001  0.001  0.059  0.059                             
## PopulatinRE 0.001  0.001  0.030  0.030  0.017                      
## PopulatinUM 0.002  0.002  0.002  0.002  0.015  0.044               
## PopulatinVA 0.017  0.015  0.001  0.001  0.002  0.044  0.058        
## PopulatinYE 0.034  0.045  0.001  0.001  0.014  0.002  0.029  0.017

p_WA_F_R_Ritchie <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_R_Ritchie", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_F_R_pop_Ritchie <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_R_pop_Ritchie", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_R_Ritchie <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R_Ritchie", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_R_pop_Ritchie <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='Ritchie')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R_pop_Ritchie", x="Population", y = "CentroidSizeRight_micrometers")+ ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

Stamenkovic-Radak Lab

WA_F_R_StamenkovicRadak_lmer <- lmer(CentroidSizeRight_micrometers ~ -1 + Population + (1|Line:Population) +
                                       (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_F,Supervisor.PI=='StamenkovicRadak')))

## boundary (singular) fit: see ?isSingular

capture.output(summary(WA_F_R_StamenkovicRadak_lmer),file = "WingArea/WA_F_R_StamenkovicRadak_sum.txt")
capture.output(anova(WA_F_R_StamenkovicRadak_lmer),file = "WingArea/WA_F_R_StamenkovicRadak_lmer.txt")
capture.output(emmeans(WA_F_R_StamenkovicRadak_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_R_StamenkovicRadak_lmer_tk.txt")

anova(WA_F_R_StamenkovicRadak_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##                Sum Sq   Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2823766639 313751849     9 114.54   40710 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_F_R_StamenkovicRadak_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeRight_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_F, Supervisor.PI == "StamenkovicRadak"))
## 
## REML criterion at convergence: 44862
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.7152 -0.5616  0.0783  0.6287  3.6999 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)    5.797  2.408  
##  Line:Population               (Intercept) 2476.853 49.768  
##  Batch                         (Intercept)    0.000  0.000  
##  Residual                                  7707.010 87.790  
## Number of obs: 3787, groups:  
## ReplicateVial:Line:Population, 124; Line:Population, 124; Batch, 5
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2775.53      13.05  113.47   212.7   <2e-16 ***
## PopulationGI  2833.73      16.56  114.81   171.2   <2e-16 ***
## PopulationKA  2813.06      13.11  115.54   214.6   <2e-16 ***
## PopulationMA  2862.04      15.79  114.99   181.2   <2e-16 ***
## PopulationMU  2850.27      13.46  112.77   211.8   <2e-16 ***
## PopulationRE  2933.26      14.52  114.76   202.0   <2e-16 ***
## PopulationUM  2886.85      14.02  115.76   205.9   <2e-16 ***
## PopulationVA  2836.46      13.51  114.43   210.0   <2e-16 ***
## PopulationYE  2823.41      13.98  114.37   202.0   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

WA_M_R_StamenkovicRadak_lmer <- lmer(CentroidSizeRight_micrometers ~ -1 + Population + (1|Line:Population) +
                                       (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_WA_M,Supervisor.PI=='StamenkovicRadak')))

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## unable to evaluate scaled gradient

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge: degenerate Hessian with 1 negative eigenvalues

## Warning: Model failed to converge with 1 negative eigenvalue: -1.4e-02

capture.output(summary(WA_M_R_StamenkovicRadak_lmer),file = "WingArea/WA_M_R_StamenkovicRadak_sum.txt")
capture.output(anova(WA_M_R_StamenkovicRadak_lmer),file = "WingArea/WA_M_R_StamenkovicRadak_lmer.txt")
capture.output(emmeans(WA_M_R_StamenkovicRadak_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "WingArea/WA_F_R_StamenkovicRadak_lmer_tk.txt")

anova(WA_M_R_StamenkovicRadak_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##                Sum Sq   Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2038283647 226475961     9 114.12   38578 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(WA_M_R_StamenkovicRadak_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## CentroidSizeRight_micrometers ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_WA_M, Supervisor.PI == "StamenkovicRadak"))
## 
## REML criterion at convergence: 42394.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -7.2672 -0.5646  0.0593  0.6360  5.9328 
## 
## Random effects:
##  Groups                        Name        Variance  Std.Dev. 
##  ReplicateVial:Line:Population (Intercept) 5.938e+00  2.436877
##  Line:Population               (Intercept) 1.970e+03 44.383290
##  Batch                         (Intercept) 9.486e-05  0.009739
##  Residual                                  5.871e+03 76.619424
## Number of obs: 3662, groups:  
## ReplicateVial:Line:Population, 124; Line:Population, 124; Batch, 5
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  2414.97      11.64  113.02   207.5   <2e-16 ***
## PopulationGI  2469.07      14.75  113.89   167.4   <2e-16 ***
## PopulationKA  2434.83      11.68  114.79   208.4   <2e-16 ***
## PopulationMA  2488.62      14.07  114.11   176.9   <2e-16 ***
## PopulationMU  2469.88      12.01  112.60   205.7   <2e-16 ***
## PopulationRE  2540.66      12.97  115.28   195.8   <2e-16 ***
## PopulationUM  2509.88      12.49  114.95   200.9   <2e-16 ***
## PopulationVA  2462.84      12.04  114.06   204.5   <2e-16 ***
## PopulationYE  2448.83      12.48  114.48   196.2   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## unable to evaluate scaled gradient
## Model failed to converge: degenerate  Hessian with 1 negative eigenvalues

p_WA_F_R_StamenkovicRadak <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_R_StamenkovicRadak", x="Population", y = "CentroidSizeRight_micrometers")+ ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_F_R_pop_StamenkovicRadak <- ggplot(data = (subset(d_WA_F,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_F_R_pop_StamenkovicRadak", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_R_StamenkovicRadak <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Batch)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R_StamenkovicRadak", x="Population", y = "CentroidSizeRight_micrometers")+ ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

p_WA_M_R_pop_StamenkovicRadak <- ggplot(data = (subset(d_WA_M,Supervisor.PI=='StamenkovicRadak')), aes(x=Population_Lat, y=CentroidSizeRight_micrometers, fill=Population)) + 
  geom_boxplot(outlier.shape = NA, notch=FALSE) + 
  labs(title="p_WA_M_R_pop_StamenkovicRadak", x="Population", y = "CentroidSizeRight_micrometers") + ylim(c(min_WAR, max_WAR))+ droseu_theme

## quartz_off_screen 
##                 2

Lab correlations

intersect_Line_WA_F  <- Reduce(intersect, list(d_WA_F$Line[d_WA_F$Supervisor.PI == "Onder"],
                                          d_WA_F$Line[d_WA_F$Supervisor.PI == "Posnien"],
                                          d_WA_F$Line[d_WA_F$Supervisor.PI == "Ritchie"],
                                          d_WA_F$Line[d_WA_F$Supervisor.PI == "StamenkovicRadak"]))
write(intersect_Line_WA_F, file ="WingArea/intersect_Line_WA_F.txt", sep = "\t")

intersect_Pop_WA_F  <- Reduce(intersect, list(d_WA_F$Population[d_WA_F$Supervisor.PI == "Onder"],
                                          d_WA_F$Population[d_WA_F$Supervisor.PI == "Posnien"],
                                          d_WA_F$Population[d_WA_F$Supervisor.PI == "Ritchie"],
                                          d_WA_F$Population[d_WA_F$Supervisor.PI == "StamenkovicRadak"]))
write(intersect_Pop_WA_F, file ="WingArea/intersect_Pop_WA_F.txt", sep = "\t")

intersect_Line_WA_M  <- Reduce(intersect, list(d_WA_M$Line[d_WA_M$Supervisor.PI == "Onder"],
                                          d_WA_M$Line[d_WA_M$Supervisor.PI == "Posnien"],
                                          d_WA_M$Line[d_WA_M$Supervisor.PI == "Ritchie"],
                                          d_WA_M$Line[d_WA_M$Supervisor.PI == "StamenkovicRadak"]))
write(intersect_Line_WA_M, file ="WingArea/intersect_Line_WA_M.txt", sep = "\t")

intersect_Pop_WA_M  <- Reduce(intersect, list(d_WA_M$Population[d_WA_M$Supervisor.PI == "Onder"],
                                          d_WA_M$Population[d_WA_M$Supervisor.PI == "Posnien"],
                                          d_WA_M$Population[d_WA_M$Supervisor.PI == "Ritchie"],
                                          d_WA_M$Population[d_WA_M$Supervisor.PI == "StamenkovicRadak"]))
write(intersect_Pop_WA_M, file ="WingArea/intersect_Pop_WA_M.txt", sep = "\t")

There are 9 populations and 5 isofemale lines for females and 9 populations and 5 isofemale lines for males have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

Left

WA_L_fitted <- combineFitted(labs = rep(c("Onder", "Posnien", "Ritchie", "StamenkovicRadak"), 2), sex = rep(c("F", "M"), each = 4), models = list(WA_F_L_Onder_lmer, WA_F_L_Posnien_lmer, WA_F_L_Ritchie_lmer, WA_F_L_StamenkovicRadak_lmer, WA_M_L_Onder_lmer, WA_M_L_Posnien_lmer, WA_M_L_Ritchie_lmer, WA_M_L_StamenkovicRadak_lmer))

p_WA_L_F_lab_cor <- function() scatterPlotMatrix(WA_L_fitted, sex = "F")
p_WA_L_M_lab_cor <- function() scatterPlotMatrix(WA_L_fitted, sex = "M")

p_WA_L_F_lab_cor()

pdf(file="WingArea/p_WA_L_F_lab_cor.pdf",width=8, height=8)
p_WA_L_F_lab_cor()
invisible(dev.off())

p_WA_L_M_lab_cor()

pdf(file="WingArea/p_WA_L_M_lab_cor.pdf",width=8, height=8)
p_WA_L_M_lab_cor()
invisible(dev.off())

Right

WA_R_fitted <- combineFitted(labs = rep(c("Onder", "Posnien", "Ritchie", "StamenkovicRadak"), 2), sex = rep(c("F", "M"), each = 4), models = list(WA_F_R_Onder_lmer, WA_F_R_Posnien_lmer, WA_F_R_Ritchie_lmer, WA_F_R_StamenkovicRadak_lmer, WA_M_R_Onder_lmer, WA_M_R_Posnien_lmer, WA_M_R_Ritchie_lmer, WA_M_R_StamenkovicRadak_lmer))

p_WA_R_F_lab_cor <- function() scatterPlotMatrix(WA_R_fitted, sex = "F")
p_WA_R_M_lab_cor <- function() scatterPlotMatrix(WA_R_fitted, sex = "M")

p_WA_R_F_lab_cor()

pdf(file="WingArea/p_WA_R_F_lab_cor.pdf",width=8, height=8)
p_WA_R_F_lab_cor()
invisible(dev.off())

p_WA_R_M_lab_cor()

pdf(file="WingArea/p_WA_R_M_lab_cor.pdf",width=8, height=8)
p_WA_R_M_lab_cor()
invisible(dev.off())

Fecundity

dir.create(file.path("Fecundity"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

#Note that the trait has been phenotyped only in females.

Contributors

Billeter Lab : Xiaocui Wang, Tiphaine Bailly, Mario Mira, Jean-Christophe Billeter

Fricke Lab : Claudia Fricke

Data Reformatting

Reading data in R

d_Fec <- read.csv("MasterSheets_Oct21_git/FEC_MasterSheet_Oct21.csv")
str(d_Fec)

## 'data.frame':    1725 obs. of  13 variables:
##  $ Supervisor.PI        : chr  "Billeter" "Billeter" "Billeter" "Billeter" ...
##  $ Diet                 : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch                : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population           : chr  "AK" "AK" "AK" "AK" ...
##  $ Line                 : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Individual           : int  1 2 3 4 5 1 2 3 4 5 ...
##  $ NumberOfAdultsEclosed: int  206 75 54 58 278 162 0 101 145 188 ...
##  $ Notes                : chr  NA NA NA NA ...
##  $ Censor               : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ Country              : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude             : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude            : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude             : int  88 88 88 88 88 88 88 88 88 88 ...

Factors need reformatting (i.e. Supervisor.PI should be coded as a factor, not character).

str(d_Fec)

## 'data.frame':    1725 obs. of  13 variables:
##  $ Supervisor.PI        : chr  "Billeter" "Billeter" "Billeter" "Billeter" ...
##  $ Diet                 : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch                : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population           : chr  "AK" "AK" "AK" "AK" ...
##  $ Line                 : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Individual           : int  1 2 3 4 5 1 2 3 4 5 ...
##  $ NumberOfAdultsEclosed: int  206 75 54 58 278 162 0 101 145 188 ...
##  $ Notes                : chr  NA NA NA NA ...
##  $ Censor               : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ Country              : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude             : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude            : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude             : int  88 88 88 88 88 88 88 88 88 88 ...

d_Fec$Supervisor.PI <- as.factor(d_Fec$Supervisor.PI)
d_Fec$Diet <- as.factor(d_Fec$Diet)
d_Fec$Batch <- as.factor(d_Fec$Batch)
d_Fec$Population <- as.factor(d_Fec$Population)
d_Fec$Population_Lat <- factor(d_Fec$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_Fec$Population_Lon <- factor(d_Fec$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_Fec$Population_Alt <- factor(d_Fec$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_Fec$Line <- as.factor(d_Fec$Line)
d_Fec$NumberOfAdultsEclosed <- as.numeric(d_Fec$NumberOfAdultsEclosed)
d_Fec$Censor <- as.factor(d_Fec$Censor)
str(d_Fec)

## 'data.frame':    1725 obs. of  16 variables:
##  $ Supervisor.PI        : Factor w/ 2 levels "Billeter","Fricke": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet                 : Factor w/ 2 levels "NS","S": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch                : Factor w/ 8 levels "1","2","3","4",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population           : Factor w/ 9 levels "AK","GI","KA",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Line                 : Factor w/ 169 levels "AK1","AK10","AK11",..: 1 1 1 1 1 11 11 11 11 11 ...
##  $ Individual           : int  1 2 3 4 5 1 2 3 4 5 ...
##  $ NumberOfAdultsEclosed: num  206 75 54 58 278 162 0 101 145 188 ...
##  $ Notes                : chr  NA NA NA NA ...
##  $ Censor               : Factor w/ 2 levels "0","1": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Country              : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude             : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude            : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude             : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat       : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon       : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt       : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_Fec <- subset(d_Fec, Censor == "0")

Now they should be in the correct format, see below.

str(d_Fec)

## 'data.frame':    1721 obs. of  16 variables:
##  $ Supervisor.PI        : Factor w/ 2 levels "Billeter","Fricke": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet                 : Factor w/ 2 levels "NS","S": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch                : Factor w/ 8 levels "1","2","3","4",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population           : Factor w/ 9 levels "AK","GI","KA",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Line                 : Factor w/ 169 levels "AK1","AK10","AK11",..: 1 1 1 1 1 11 11 11 11 11 ...
##  $ Individual           : int  1 2 3 4 5 1 2 3 4 5 ...
##  $ NumberOfAdultsEclosed: num  206 75 54 58 278 162 0 101 145 188 ...
##  $ Notes                : chr  NA NA NA NA ...
##  $ Censor               : Factor w/ 2 levels "0","1": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Country              : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude             : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude            : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude             : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat       : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon       : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt       : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

Descriptive statistics for fecundity at the line level, without batch information :

Descriptive statistics at the population level, with batch information :

Descriptive statistics at the population level, without batch information :

Plots and Linear Models by Lab

Overview

min_Fec <- min(d_Fec$NumberOfAdultsEclosed)
max_Fec <- max(d_Fec$NumberOfAdultsEclosed)

y-axis is scaled by the minimum (0) and maximum (306) values in the full data set.

p_Fec_all <- ggplot(data = d_Fec, aes(x=Population_Lat, y=NumberOfAdultsEclosed, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Fec_all", x="Population", y = "DW_micrograms") + ylim(c(min_Fec, max_Fec))+ droseu_theme

## quartz_off_screen 
##                 2

Billeter Lab

Fec_lmer_Billeter <- lmer(NumberOfAdultsEclosed ~ -1 + Population + (1|Population:Line), data = d_Fec[d_Fec$Supervisor.PI == "Billeter",]) 
capture.output(anova(Fec_lmer_Billeter),file = "Fecundity/Fec_lmer_Billeter.txt")
capture.output(summary(Fec_lmer_Billeter),file = "Fecundity/Fec_lmer_Billeter_sum.txt")
capture.output(emmeans(Fec_lmer_Billeter, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"), file = "Fecundity/Fec_lmer_Billeter_tk.txt")

model_means_Fec_Billeter <- emmeans(object = Fec_lmer_Billeter, specs = ~ Population)
model_means_Fec_Billeter_cld <- cld(object = model_means_Fec_Billeter, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(Fec_lmer_Billeter)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 2150407  238934     9 148.4  86.477 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Fec_lmer_Billeter)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: NumberOfAdultsEclosed ~ -1 + Population + (1 | Population:Line)
##    Data: d_Fec[d_Fec$Supervisor.PI == "Billeter", ]
## 
## REML criterion at convergence: 8773.2
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -3.02323 -0.55643 -0.03579  0.53645  2.97008 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Population:Line (Intercept) 1113     33.36   
##  Residual                    2763     52.56   
## Number of obs: 805, groups:  Population:Line, 160
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   97.067      9.154 149.480  10.603  < 2e-16 ***
## PopulationGI   74.193     10.465 144.161   7.090 5.54e-11 ***
## PopulationKA  109.970      9.128 148.291  12.047  < 2e-16 ***
## PopulationMA   99.120      9.868 146.604  10.045  < 2e-16 ***
## PopulationMU   93.842      9.546 155.094   9.830  < 2e-16 ***
## PopulationRE   67.854     10.896 147.650   6.227 4.68e-09 ***
## PopulationUM  102.827     10.537 148.601   9.758  < 2e-16 ***
## PopulationVA   86.333      9.080 145.006   9.508  < 2e-16 ***
## PopulationYE   63.937      9.179 151.258   6.966 9.36e-11 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Fec_Billeter <- ggplot(data = (subset(d_Fec,Supervisor.PI=='Billeter')), aes(x=Population_Lat, y=NumberOfAdultsEclosed, fill=Batch)) +  geom_boxplot(outlier.shape = NA, notch=FALSE) + labs(title="p_Fec_Billeter", x="Population", y = "NumberOfAdultsEclosed") + ylim(c(min_Fec, max_Fec))+ droseu_theme

## quartz_off_screen 
##                 2

p_Fec_pop_Billeter <- ggplot(data = (subset(d_Fec,Supervisor.PI=='Billeter')), aes(x=Population_Lat, y=NumberOfAdultsEclosed, fill=Population)) + geom_boxplot(outlier.shape = NA, notch=FALSE) + labs(title="p_Fec_pop_Billeter", x="Population", y = "NumberOfAdultsEclosed") + ylim(c(min_Fec, max_Fec))+geom_text(data = model_means_Fec_Billeter_cld, aes(x = Population, label = cld), y = 1, vjust = 1) +  droseu_theme

## quartz_off_screen 
##                 2

Fricke Lab

model_means_Fec_Fricke <- emmeans(object = Fec_lmer_Fricke, specs = ~ Population)
model_means_Fec_Fricke_cld <- cld(object = model_means_Fec_Fricke, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(Fec_lmer_Fricke)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 346953   38550     9 54.703  30.968 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Fec_lmer_Fricke)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: NumberOfAdultsEclosed ~ -1 + Population + (1 | Population:Line) +  
##     (1 | Batch)
##    Data: d_Fec[d_Fec$Supervisor.PI == "Fricke", ]
## 
## REML criterion at convergence: 9249.1
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.8102 -0.6481 -0.0705  0.5894  3.4523 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Population:Line (Intercept)  455.06  21.332  
##  Batch           (Intercept)   41.26   6.423  
##  Residual                    1244.86  35.283  
## Number of obs: 916, groups:  Population:Line, 161; Batch, 8
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   61.074      6.779  95.724   9.009 2.05e-14 ***
## PopulationGI   50.491      7.392 108.616   6.830 5.10e-10 ***
## PopulationKA   50.776      6.613  85.361   7.678 2.45e-11 ***
## PopulationMA   55.985      6.500  82.745   8.613 3.97e-13 ***
## PopulationMU   56.867      6.667  93.038   8.529 2.62e-13 ***
## PopulationRE   57.462      7.300 105.927   7.871 3.20e-12 ***
## PopulationUM   46.999      6.848  93.859   6.863 7.16e-10 ***
## PopulationVA   53.882      6.631  93.870   8.126 1.75e-12 ***
## PopulationYE   52.842      6.608  91.405   7.996 3.80e-12 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.127                                                  
## PopulatinKA 0.143  0.133                                           
## PopulatinMA 0.147  0.134  0.154                                    
## PopulatinMU 0.129  0.116  0.136  0.134                             
## PopulatinRE 0.129  0.119  0.136  0.143  0.117                      
## PopulatinUM 0.135  0.124  0.140  0.152  0.121  0.132               
## PopulatinVA 0.134  0.121  0.141  0.150  0.131  0.127  0.134        
## PopulatinYE 0.137  0.125  0.143  0.151  0.127  0.132  0.140  0.134

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

Lab correlations

intersect_Line_Fec <- intersect(d_Fec$Line[d_Fec$Supervisor.PI == "Billeter"], 
                           d_Fec$Line[d_Fec$Supervisor.PI == "Fricke"])
write(intersect_Line_Fec, file ="Fecundity/intersect_Line_Fec.txt", sep = "\t")


intersect_Pop_Fec <- intersect(d_Fec$Line[d_Fec$Supervisor.PI == "Billeter"], 
                           d_Fec$Line[d_Fec$Supervisor.PI == "Fricke"])
write(intersect_Pop_Fec, file ="Fecundity/intersect_Pop_Fec.txt", sep = "\t")

There are 152 populations and 152 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

Fec_fitted <- combineFitted(labs = c("Billeter", "Fricke"), sex = rep("NA", 2), models = list(Fec_lmer_Billeter, Fec_lmer_Fricke))

p_Fec_lab_cor <- function() scatterPlotMatrix(Fec_fitted, sex = "NA")

p_Fec_lab_cor()

pdf(file="Fecundity/p_Fec_lab_cor.pdf",width=8, height=8)
p_Fec_lab_cor()
invisible(dev.off())

Lifespan

dir.create(file.path("Lifespan"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Flatt Lab : Esra Durmaz, Envel Kerdaffrec, Thibault Schowing, Virginie Thieu, Marisa Rodrigues, Thomas Flatt

Parsch Lab : Amanda Glaser-Schmitt, Eliza Argyridou, John Parsch

Pasyukova Lab : Natalia Roshina, Alexander Symonenko, Mikhail Trostnikov, Evgenia Tsybul’ko, Ekaterina Veselkina, Olga Rybina, Elena Pasyukova

Data Reformatting

d_LS_L <- read.csv("MasterSheets_Oct21_git/LSL_MasterSheet_Oct21.csv")

str(d_LS_L)

## 'data.frame':    39844 obs. of  15 variables:
##  $ Supervisor.PI      : chr  "Pasyukova" "Pasyukova" "Pasyukova" "Pasyukova" ...
##  $ Diet               : chr  "S" "S" "S" "S" ...
##  $ Batch              : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population         : chr  "AK" "AK" "AK" "AK" ...
##  $ Line               : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Sex                : chr  "F" "F" "F" "F" ...
##  $ ReplicateVialOld   : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial      : chr  "Pasyukova_1_AK1_1" "Pasyukova_1_AK1_1" "Pasyukova_1_AK1_1" "Pasyukova_1_AK1_1" ...
##  $ Individual         : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ Censor             : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ LSL_AgeAtDeath_days: int  10 37 41 42 42 45 45 45 45 46 ...
##  $ Country            : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude           : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude          : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude           : int  88 88 88 88 88 88 88 88 88 88 ...

d_LS_L$Supervisor.PI <- as.factor(d_LS_L$Supervisor.PI)
d_LS_L$Diet <- as.factor(d_LS_L$Diet)
d_LS_L$Batch <- as.factor(d_LS_L$Batch)
d_LS_L$Population <- as.factor(d_LS_L$Population)
d_LS_L$Population_Lat <- factor(d_LS_L$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_LS_L$Population_Lon <- factor(d_LS_L$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_LS_L$Population_Alt <- factor(d_LS_L$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_LS_L$Line <- as.factor(d_LS_L$Line)
d_LS_L$ReplicateVial <- as.factor(d_LS_L$ReplicateVial)
d_LS_L$LSL_AgeAtDeath_days <- as.numeric(d_LS_L$LSL_AgeAtDeath_days)
d_LS_L$Censor <- as.factor(d_LS_L$Censor)

str(d_LS_L)

## 'data.frame':    39844 obs. of  18 variables:
##  $ Supervisor.PI      : Factor w/ 2 levels "Parsch","Pasyukova": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Diet               : Factor w/ 2 levels "NS","S": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Batch              : Factor w/ 2 levels "1","2": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population         : Factor w/ 9 levels "AK","GI","KA",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Line               : Factor w/ 171 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex                : chr  "F" "F" "F" "F" ...
##  $ ReplicateVialOld   : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial      : Factor w/ 2032 levels "Parsch_1_AK1_1",..: 681 681 681 681 681 681 681 681 681 681 ...
##  $ Individual         : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ Censor             : Factor w/ 2 levels "0","1": 1 1 1 1 1 1 1 1 1 1 ...
##  $ LSL_AgeAtDeath_days: num  10 37 41 42 42 45 45 45 45 46 ...
##  $ Country            : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude           : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude          : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude           : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat     : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon     : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt     : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_LS_L_F <-subset(d_LS_L,Sex=='F')
d_LS_L_M <-subset(d_LS_L,Sex=='M')

d_LS_P <- read.csv("MasterSheets_Oct21_git/LSP_MasterSheet_Oct21.csv")

str(d_LS_P)

## 'data.frame':    10612 obs. of  14 variables:
##  $ Supervisor.PI      : chr  "Flatt" "Flatt" "Flatt" "Flatt" ...
##  $ Diet               : chr  "S" "S" "S" "S" ...
##  $ Batch              : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population         : chr  "AK" "AK" "AK" "AK" ...
##  $ Sex                : chr  "F" "F" "F" "F" ...
##  $ ReplicateCageOld   : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateCage      : chr  "Flatt_AK_F_1" "Flatt_AK_F_1" "Flatt_AK_F_1" "Flatt_AK_F_1" ...
##  $ Individual         : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ Censor             : int  1 1 1 1 0 0 0 1 0 0 ...
##  $ LSP_AgeAtDeath_days: int  11 11 17 23 26 26 33 33 40 43 ...
##  $ Country            : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude           : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude          : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude           : int  88 88 88 88 88 88 88 88 88 88 ...

d_LS_P$Supervisor.PI <- as.factor(d_LS_P$Supervisor.PI)
d_LS_P$Diet <- as.factor(d_LS_P$Diet)
d_LS_P$Batch <- as.factor(d_LS_P$Batch)
d_LS_P$Population <- as.factor(d_LS_P$Population)
d_LS_P$Population_Lat <- factor(d_LS_P$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_LS_P$Population_Lon <- factor(d_LS_P$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_LS_P$Population_Alt <- factor(d_LS_P$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_LS_P$ReplicateCage <- as.factor(d_LS_P$ReplicateCage)
d_LS_P$LSP_AgeAtDeath_days <- as.numeric(d_LS_P$LSP_AgeAtDeath_days)
d_LS_P$Censor <- as.factor(d_LS_P$Censor)

str(d_LS_P)

## 'data.frame':    10612 obs. of  17 variables:
##  $ Supervisor.PI      : Factor w/ 1 level "Flatt": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet               : Factor w/ 1 level "S": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch              : Factor w/ 1 level "1": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population         : Factor w/ 9 levels "AK","GI","KA",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex                : chr  "F" "F" "F" "F" ...
##  $ ReplicateCageOld   : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateCage      : Factor w/ 180 levels "Flatt_AK_F_1",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Individual         : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ Censor             : Factor w/ 2 levels "0","1": 2 2 2 2 1 1 1 2 1 1 ...
##  $ LSP_AgeAtDeath_days: num  11 11 17 23 26 26 33 33 40 43 ...
##  $ Country            : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude           : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude          : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude           : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat     : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon     : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt     : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_LS_P_F <-subset(d_LS_P,Sex=='F')
d_LS_P_M <-subset(d_LS_P,Sex=='M')

d_LS_M <- read.csv("MasterSheets_Oct21_git/LSM_MasterSheet_Oct21.csv")

str(d_LS_M)

## 'data.frame':    50456 obs. of  12 variables:
##  $ Supervisor.PI      : chr  "Pasyukova" "Pasyukova" "Pasyukova" "Pasyukova" ...
##  $ Diet               : chr  "S" "S" "S" "S" ...
##  $ Batch              : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population         : chr  "AK" "AK" "AK" "AK" ...
##  $ Sex                : chr  "F" "F" "F" "F" ...
##  $ Individual         : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ Censor             : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ LSM_AgeAtDeath_days: int  10 37 41 42 42 45 45 45 45 46 ...
##  $ Country            : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude           : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude          : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude           : int  88 88 88 88 88 88 88 88 88 88 ...

d_LS_M$Supervisor.PI <- as.factor(d_LS_M$Supervisor.PI)
d_LS_M$Diet <- as.factor(d_LS_M$Diet)
d_LS_M$Batch <- as.factor(d_LS_M$Batch)
d_LS_M$Population <- as.factor(d_LS_M$Population)
d_LS_M$Population_Lat <- factor(d_LS_M$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_LS_M$Population_Lon <- factor(d_LS_M$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_LS_M$Population_Alt <- factor(d_LS_M$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_LS_M$LSM_AgeAtDeath_days <- as.numeric(d_LS_M$LSM_AgeAtDeath_days)
d_LS_M$Censor <- as.factor(d_LS_M$Censor)

str(d_LS_M)

## 'data.frame':    50456 obs. of  15 variables:
##  $ Supervisor.PI      : Factor w/ 3 levels "Flatt","Parsch",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ Diet               : Factor w/ 2 levels "NS","S": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Batch              : Factor w/ 2 levels "1","2": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population         : Factor w/ 9 levels "AK","GI","KA",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex                : chr  "F" "F" "F" "F" ...
##  $ Individual         : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ Censor             : Factor w/ 2 levels "0","1": 1 1 1 1 1 1 1 1 1 1 ...
##  $ LSM_AgeAtDeath_days: num  10 37 41 42 42 45 45 45 45 46 ...
##  $ Country            : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude           : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude          : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude           : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat     : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon     : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt     : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_LS_M_F <-subset(d_LS_M,Sex=='F')
d_LS_M_M <-subset(d_LS_M,Sex=='M')

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

table_LS_L_Line_wbatch <- write.csv(d_LS_L %>% group_by(Supervisor.PI, Batch, Sex, Population, Line) %>% 
                                    summarise_at(vars(LSL_AgeAtDeath_days), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "Lifespan/table_LS_L_Line_wbatch.csv", row.names = T)
table_LS_L_Line_wbatch <- read.csv("Lifespan/table_LS_L_Line_wbatch.csv")

Descriptive statistics at the line level, without batch information :

table_LS_L_Line_wobatch <- write.csv(d_LS_L %>% group_by(Supervisor.PI, Sex, Population, Line) %>% 
                                    summarise_at(vars(LSL_AgeAtDeath_days), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "Lifespan/table_LS_L_Line_wobatch.csv", row.names = T)
table_LS_L_Line_wobatch <- read.csv("Lifespan/table_LS_L_Line_wobatch.csv")

Descriptive statistics at the population level, with batch information :

table_LS_M_Pop_wbatch <- write.csv(d_LS_M %>% group_by(Supervisor.PI, Batch, Sex, Population) %>% 
                                   summarise_at(vars(LSM_AgeAtDeath_days), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Lifespan/table_LS_M_Pop_wbatch.csv", row.names = T)

table_LS_M_Pop_wbatch <- read.csv("Lifespan/table_LS_M_Pop_wbatch.csv")

Descriptive statistics at the population level, without batch information :

table_LS_M_Pop_wobatch <- write.csv(d_LS_M %>% group_by(Supervisor.PI, Sex, Population) %>% 
                                   summarise_at(vars(LSM_AgeAtDeath_days), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Lifespan/table_LS_M_Pop_wobatch.csv", row.names = T)
table_LS_M_Pop_wobatch <- read.csv("Lifespan/table_LS_M_Pop_wobatch.csv")

Plots and Linear Models by Lab

Overview

min_LS <- min(d_LS_M$LSM_AgeAtDeath_days)
max_LS <- max(d_LS_M$LSM_AgeAtDeath_days)

y-axis is scaled by the minimum (1) and maximum (112) values in the full data set.

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

Flatt Lab

LS_P_F_Flatt_lmer <- lmer(LSP_AgeAtDeath_days ~ -1 + Population + (1|Population:ReplicateCage), data = filter(d_LS_P_F, Censor == "0" & Supervisor.PI == "Flatt"))
capture.output(summary(LS_P_F_Flatt_lmer),file = "Lifespan/LS_P_F_Flatt_lmer.txt")
capture.output(anova(LS_P_F_Flatt_lmer),file = "Lifespan/LS_P_F_Flatt_lmer.txt")
capture.output(emmeans(LS_P_F_Flatt_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Lifespan/LS_P_F_Flatt_lmer_tk.txt")

anova(LS_P_F_Flatt_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 4720794  524533     9 80.872  3341.1 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LS_P_F_Flatt_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: LSP_AgeAtDeath_days ~ -1 + Population + (1 | Population:ReplicateCage)
##    Data: filter(d_LS_P_F, Censor == "0" & Supervisor.PI == "Flatt")
## 
## REML criterion at convergence: 36116.1
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.3943 -0.6327 -0.0087  0.6402  3.5292 
## 
## Random effects:
##  Groups                   Name        Variance Std.Dev.
##  Population:ReplicateCage (Intercept)   5.11    2.261  
##  Residual                             156.99   12.530  
## Number of obs: 4567, groups:  Population:ReplicateCage, 90
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  60.0334     0.9132 83.3744   65.74   <2e-16 ***
## PopulationGI  42.9477     0.8945 76.7627   48.01   <2e-16 ***
## PopulationKA  55.4246     0.8972 77.6946   61.77   <2e-16 ***
## PopulationMA  50.5195     0.9074 81.2343   55.68   <2e-16 ***
## PopulationMU  57.4202     0.9103 82.2518   63.08   <2e-16 ***
## PopulationRE  45.6074     0.9038 80.0209   50.46   <2e-16 ***
## PopulationUM  51.6708     0.9208 85.8668   56.12   <2e-16 ***
## PopulationVA  56.2711     0.9109 82.3980   61.77   <2e-16 ***
## PopulationYE  49.6624     0.9011 79.0852   55.11   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

LS_P_M_Flatt_lmer <- lmer(LSP_AgeAtDeath_days ~ -1 + Population + (1|Population:ReplicateCage), data = filter(d_LS_P_M, Censor == "0" & Supervisor.PI == "Flatt"))
capture.output(summary(LS_P_M_Flatt_lmer),file = "Lifespan/LS_P_M_Flatt_lmer_sum.txt")
capture.output(anova(LS_P_M_Flatt_lmer),file = "Lifespan/LS_P_M_Flatt_lmer.txt")
capture.output(emmeans(LS_P_M_Flatt_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Lifespan/LS_P_M_Flatt_lmer_tk.txt")

anova(LS_P_M_Flatt_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 5928693  658744     9 80.244  3768.4 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LS_P_M_Flatt_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: LSP_AgeAtDeath_days ~ -1 + Population + (1 | Population:ReplicateCage)
##    Data: filter(d_LS_P_M, Censor == "0" & Supervisor.PI == "Flatt")
## 
## REML criterion at convergence: 35650
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.9146 -0.6656 -0.0467  0.6755  3.0925 
## 
## Random effects:
##  Groups                   Name        Variance Std.Dev.
##  Population:ReplicateCage (Intercept)   3.827   1.956  
##  Residual                             174.806  13.221  
## Number of obs: 4450, groups:  Population:ReplicateCage, 90
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  51.6194     0.8739 86.0482   59.06   <2e-16 ***
## PopulationGI  50.8448     0.8507 77.1614   59.77   <2e-16 ***
## PopulationKA  56.0643     0.8543 78.6017   65.63   <2e-16 ***
## PopulationMA  53.0393     0.8502 77.1549   62.38   <2e-16 ***
## PopulationMU  55.6448     0.8667 83.3740   64.20   <2e-16 ***
## PopulationRE  50.4210     0.8510 77.2981   59.25   <2e-16 ***
## PopulationUM  52.3560     0.8536 78.1773   61.34   <2e-16 ***
## PopulationVA  52.0358     0.8796 87.9588   59.16   <2e-16 ***
## PopulationYE  52.3283     0.8531 78.1557   61.34   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

Parsch Lab

LS_L_F_Parsch_lmer <- lmer(LSL_AgeAtDeath_days ~ -1 + Population + (1|Batch) + (1|Population:Line) + (1|Line:ReplicateVial), data = filter(d_LS_L_F, Censor == "0" & Supervisor.PI == "Parsch"))
capture.output(summary(LS_L_F_Parsch_lmer),file = "Lifespan/LS_L_F_Parsch_lmer_sum.txt")
capture.output(anova(LS_L_F_Parsch_lmer),file = "Lifespan/LS_L_F_Parsch_lmer.txt")
capture.output(emmeans(LS_L_F_Parsch_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Lifespan/LS_L_F_Parsch_lmer_tk.txt")

anova(LS_L_F_Parsch_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value  Pr(>F)  
## Population 130058   14451     9     2  54.048 0.01829 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LS_L_F_Parsch_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## LSL_AgeAtDeath_days ~ -1 + Population + (1 | Batch) + (1 | Population:Line) +  
##     (1 | Line:ReplicateVial)
##    Data: filter(d_LS_L_F, Censor == "0" & Supervisor.PI == "Parsch")
## 
## REML criterion at convergence: 42206.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.6820 -0.5293  0.1220  0.6479  2.5964 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  Line:ReplicateVial (Intercept)  46.61    6.827  
##  Population:Line    (Intercept)  21.52    4.639  
##  Batch              (Intercept)  12.92    3.594  
##  Residual                       267.37   16.351  
## Number of obs: 4934, groups:  
## Line:ReplicateVial, 679; Population:Line, 135; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)   
## PopulationAK   59.625      3.000  1.861   19.88  0.00350 **
## PopulationGI   54.909      3.006  1.877   18.27  0.00395 **
## PopulationKA   56.866      3.003  1.871   18.93  0.00374 **
## PopulationMA   47.306      3.011  1.889   15.71  0.00509 **
## PopulationMU   54.014      3.003  1.871   17.98  0.00412 **
## PopulationRE   46.438      3.001  1.866   15.47  0.00551 **
## PopulationUM   49.083      3.004  1.872   16.34  0.00491 **
## PopulationVA   56.420      2.997  1.855   18.83  0.00393 **
## PopulationYE   54.592      3.007  1.881   18.15  0.00396 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.717                                                  
## PopulatinKA 0.717  0.715                                           
## PopulatinMA 0.715  0.713  0.715                                    
## PopulatinMU 0.717  0.715  0.716  0.714                             
## PopulatinRE 0.717  0.716  0.717  0.715  0.717                      
## PopulatinUM 0.717  0.715  0.716  0.714  0.716  0.717               
## PopulatinVA 0.718  0.717  0.718  0.716  0.718  0.718  0.718        
## PopulatinYE 0.716  0.714  0.715  0.714  0.715  0.716  0.715  0.717

LS_L_M_Parsch_lmer <- lmer(LSL_AgeAtDeath_days ~ -1 + Population + (1|Batch) + (1|Population:Line) + (1|Line:ReplicateVial), data = filter(d_LS_L_M, Censor == "0" & Supervisor.PI == "Parsch"))
capture.output(summary(LS_L_M_Parsch_lmer),file = "Lifespan/LS_L_M_Parsch_lmer_sum.txt")
capture.output(anova(LS_L_M_Parsch_lmer),file = "Lifespan/LS_L_M_Parsch_lmer.txt")
capture.output(emmeans(LS_L_M_Parsch_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Lifespan/LS_L_M_Parsch_lmer_tk.txt")

anova(LS_L_M_Parsch_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value   Pr(>F)   
## Population 725146   80572     9     2  226.58 0.004402 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LS_L_M_Parsch_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## LSL_AgeAtDeath_days ~ -1 + Population + (1 | Batch) + (1 | Population:Line) +  
##     (1 | Line:ReplicateVial)
##    Data: filter(d_LS_L_M, Censor == "0" & Supervisor.PI == "Parsch")
## 
## REML criterion at convergence: 42301.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.2393 -0.5900  0.0897  0.6900  2.7320 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  Line:ReplicateVial (Intercept)  13.150   3.626  
##  Population:Line    (Intercept)  34.973   5.914  
##  Batch              (Intercept)   2.336   1.528  
##  Residual                       355.597  18.857  
## Number of obs: 4821, groups:  
## Line:ReplicateVial, 679; Population:Line, 135; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   60.473      2.089 11.236   28.94 6.57e-12 ***
## PopulationGI   51.369      2.087 11.190   24.61 4.25e-11 ***
## PopulationKA   56.795      2.087 11.183   27.21 1.42e-11 ***
## PopulationMA   50.958      2.089 11.228   24.39 4.42e-11 ***
## PopulationMU   58.601      2.080 11.036   28.18 1.24e-11 ***
## PopulationRE   49.221      2.081 11.065   23.65 7.97e-11 ***
## PopulationUM   56.873      2.089 11.226   27.23 1.32e-11 ***
## PopulationVA   59.741      2.086 11.157   28.64 8.42e-12 ***
## PopulationYE   51.676      2.088 11.202   24.75 3.92e-11 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.268                                                  
## PopulatinKA 0.267  0.268                                           
## PopulatinMA 0.267  0.267  0.269                                    
## PopulatinMU 0.269  0.269  0.269  0.269                             
## PopulatinRE 0.268  0.268  0.269  0.269  0.270                      
## PopulatinUM 0.267  0.268  0.268  0.268  0.269  0.269               
## PopulatinVA 0.268  0.268  0.268  0.268  0.269  0.269  0.268        
## PopulatinYE 0.267  0.268  0.269  0.268  0.269  0.269  0.268  0.268

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

Pasyukova Lab

LS_L_F_Pasyukova_lmer <- lmer(LSL_AgeAtDeath_days ~ -1 + Population + (1|Batch) + (1|Population:Line) + (1|Line:ReplicateVial), data = filter(d_LS_L_F, Censor == "0" & Supervisor.PI == "Pasyukova"))
capture.output(summary(LS_L_F_Pasyukova_lmer),file = "Lifespan/LS_L_F_Pasyukova_lmer_sum.txt")
capture.output(anova(LS_L_F_Pasyukova_lmer),file = "Lifespan/LS_L_F_Pasyukova_lmer.txt")
capture.output(emmeans(LS_L_F_Pasyukova_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Lifespan/LS_L_F_Pasyukova_lmer_tk.txt")

anova(LS_L_F_Pasyukova_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value   Pr(>F)   
## Population 263591   29288     9     2  139.58 0.007133 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LS_L_F_Pasyukova_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## LSL_AgeAtDeath_days ~ -1 + Population + (1 | Batch) + (1 | Population:Line) +  
##     (1 | Line:ReplicateVial)
##    Data: filter(d_LS_L_F, Censor == "0" & Supervisor.PI == "Pasyukova")
## 
## REML criterion at convergence: 112305.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.4209 -0.5713  0.0810  0.6441  3.8348 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  Line:ReplicateVial (Intercept)  39.803   6.309  
##  Population:Line    (Intercept)  24.787   4.979  
##  Batch              (Intercept)   2.667   1.633  
##  Residual                       209.826  14.485  
## Number of obs: 13520, groups:  
## Line:ReplicateVial, 1352; Population:Line, 169; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   43.977      1.676  4.106   26.24 9.95e-06 ***
## PopulationGI   43.844      1.956  7.527   22.41 3.64e-08 ***
## PopulationKA   46.763      1.719  4.541   27.20 3.33e-06 ***
## PopulationMA   43.617      1.719  4.541   25.37 4.56e-06 ***
## PopulationMU   43.921      1.719  4.541   25.55 4.42e-06 ***
## PopulationRE   44.179      1.833  5.842   24.10 4.51e-07 ***
## PopulationUM   39.324      1.771  5.100   22.21 2.85e-06 ***
## PopulationVA   44.056      1.719  4.541   25.62 4.36e-06 ***
## PopulationYE   41.305      1.719  4.541   24.02 5.83e-06 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.408                                                  
## PopulatinKA 0.464  0.397                                           
## PopulatinMA 0.464  0.397  0.452                                    
## PopulatinMU 0.464  0.397  0.452  0.452                             
## PopulatinRE 0.435  0.373  0.424  0.424  0.424                      
## PopulatinUM 0.450  0.386  0.439  0.439  0.439  0.412               
## PopulatinVA 0.464  0.397  0.452  0.452  0.452  0.424  0.439        
## PopulatinYE 0.464  0.397  0.452  0.452  0.452  0.424  0.439  0.452

LS_L_M_Pasyukova_lmer <- lmer(LSL_AgeAtDeath_days ~ -1 + Population + (1|Batch) + (1|Population:Line) + (1|Line:ReplicateVial), data = filter(d_LS_L_M, Censor == "0" & Supervisor.PI == "Pasyukova"))

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00449967 (tol = 0.002, component 1)

capture.output(summary(LS_L_M_Pasyukova_lmer),file = "Lifespan/LS_L_M_Pasyukova_lmer_sum.txt")
capture.output(anova(LS_L_M_Pasyukova_lmer),file = "Lifespan/LS_L_M_Pasyukova_lmer.txt")
capture.output(emmeans(LS_L_M_Pasyukova_lmer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Lifespan/LS_L_M_Pasyukova_lmer_tk.txt")

anova(LS_L_M_Pasyukova_lmer)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value  Pr(>F)   
## Population 730175   81131     9     2  290.93 0.00343 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LS_L_M_Pasyukova_lmer)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## LSL_AgeAtDeath_days ~ -1 + Population + (1 | Batch) + (1 | Population:Line) +  
##     (1 | Line:ReplicateVial)
##    Data: filter(d_LS_L_M, Censor == "0" & Supervisor.PI == "Pasyukova")
## 
## REML criterion at convergence: 116093.1
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.0485 -0.5537  0.0609  0.6453  2.8822 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  Line:ReplicateVial (Intercept)  49.623   7.044  
##  Population:Line    (Intercept)  31.736   5.633  
##  Batch              (Intercept)   1.325   1.151  
##  Residual                       278.870  16.699  
## Number of obs: 13520, groups:  
## Line:ReplicateVial, 1352; Population:Line, 169; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   50.476      1.597 11.820   31.62 8.68e-13 ***
## PopulationGI   47.763      1.963 25.115   24.33  < 2e-16 ***
## PopulationKA   53.509      1.654 13.517   32.34 3.38e-14 ***
## PopulationMA   47.559      1.654 13.517   28.75 1.63e-13 ***
## PopulationMU   50.604      1.654 13.517   30.59 7.12e-14 ***
## PopulationRE   48.307      1.804 18.608   26.77 2.56e-16 ***
## PopulationUM   43.614      1.723 15.707   25.32 3.70e-14 ***
## PopulationVA   49.283      1.654 13.517   29.79 1.01e-13 ***
## PopulationYE   44.476      1.654 13.517   26.88 3.97e-13 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.213                                                  
## PopulatinKA 0.252  0.205                                           
## PopulatinMA 0.252  0.205  0.243                                    
## PopulatinMU 0.252  0.205  0.243  0.243                             
## PopulatinRE 0.231  0.188  0.223  0.223  0.223                      
## PopulatinUM 0.242  0.197  0.234  0.234  0.234  0.214               
## PopulatinVA 0.252  0.205  0.243  0.243  0.243  0.223  0.234        
## PopulatinYE 0.252  0.205  0.243  0.243  0.243  0.223  0.234  0.243 
## optimizer (nloptwrap) convergence code: 0 (OK)
## Model failed to converge with max|grad| = 0.00449967 (tol = 0.002, component 1)

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

## quartz_off_screen 
##                 2

Survival

d_LS_P_surv <- d_LS_P %>% mutate(Censor = ifelse(Censor == 0, 1, 0))
d_LS_L_surv <- d_LS_L %>% mutate(Censor = ifelse(Censor == 0, 1, 0))

Flatt Lab

Females

LS_F_Flatt_survfit <- survfit(Surv(LSP_AgeAtDeath_days) ~ Population, data = filter(d_LS_P_surv, Supervisor.PI == "Flatt", Sex == "F"))

p_LS_F_Flatt_survfit <- ggsurvplot(LS_F_Flatt_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "LS_F_Flatt_survfit")

p_LS_F_Flatt_survfit$plot

pdf(file="Lifespan/p_LS_F_Flatt_survfit.pdf", width=8, height=5)
p_LS_F_Flatt_survfit$plot
invisible(dev.off())

Males

LS_M_Flatt_survfit <- survfit(Surv(LSP_AgeAtDeath_days) ~ Population, data = filter(d_LS_P_surv, Supervisor.PI == "Flatt", Sex == "M"))

p_LS_M_Flatt_survfit <- ggsurvplot(LS_M_Flatt_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "LS_M_Flatt_survfit")

p_LS_M_Flatt_survfit$plot

pdf(file="Lifespan/p_LS_M_Flatt_survfit.pdf", width=8, height=5)
p_LS_M_Flatt_survfit$plot
invisible(dev.off())

Parsch Lab

Females

LS_F_Parsch_survfit <- survfit(Surv(LSL_AgeAtDeath_days) ~ Population, data = filter(d_LS_L_surv, Supervisor.PI == "Parsch", Sex == "F"))

p_LS_F_Parsch_survfit <- ggsurvplot(LS_F_Parsch_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "LS_F_Parsch_survfit")

p_LS_F_Parsch_survfit$plot

pdf(file="Lifespan/p_LS_F_Parsch_survfit.pdf", width=8, height=5)
p_LS_F_Parsch_survfit$plot
invisible(dev.off())

Males

LS_M_Parsch_survfit <- survfit(Surv(LSL_AgeAtDeath_days) ~ Population, data = filter(d_LS_L_surv, Supervisor.PI == "Parsch", Sex == "M"))

p_LS_M_Parsch_survfit <- ggsurvplot(LS_M_Parsch_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "LS_M_Parsch_survfit")

p_LS_M_Parsch_survfit$plot

pdf(file="Lifespan/p_LS_M_Parsch_survfit.pdf", width=8, height=5)
p_LS_M_Parsch_survfit$plot
invisible(dev.off())

Pasyukova Lab

Females

LS_F_Pasyukova_survfit <- survfit(Surv(LSL_AgeAtDeath_days) ~ Population, data = filter(d_LS_L_surv, Supervisor.PI == "Pasyukova", Sex == "F"))

p_LS_F_Pasyukova_survfit <- ggsurvplot(LS_F_Pasyukova_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "LS_F_Pasyukova_survfit")

p_LS_F_Pasyukova_survfit$plot

pdf(file="Lifespan/p_LS_F_Pasyukova_survfit.pdf", width=8, height=5)
p_LS_F_Pasyukova_survfit$plot
invisible(dev.off())

Males

LS_M_Pasyukova_survfit <- survfit(Surv(LSL_AgeAtDeath_days) ~ Population, data = filter(d_LS_L_surv, Supervisor.PI == "Pasyukova", Sex == "M"))

p_LS_M_Pasyukova_survfit <- ggsurvplot(LS_M_Pasyukova_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "LS_M_Pasyukova_survfit")

p_LS_M_Pasyukova_survfit$plot

pdf(file="Lifespan/p_LS_M_Pasyukova_survfit.pdf", width=8, height=5)
p_LS_M_Pasyukova_survfit$plot
invisible(dev.off())

coxme

LS_P_F_coxme_Flatt <- coxme(Surv(LSP_AgeAtDeath_days, Censor) ~ Population + (1|Population/ReplicateCage), data = filter(d_LS_P_surv, Supervisor.PI == "Flatt", Sex == "F"))
capture.output(summary(LS_P_F_coxme_Flatt), file = "Starvation/LS_P_F_coxme_Flatt_sum.txt")
capture.output(anova(LS_P_F_coxme_Flatt), file = "Starvation/LS_P_F_coxme_Flatt.txt")

LS_P_M_coxme_Flatt <- coxme(Surv(LSP_AgeAtDeath_days, Censor) ~ Population + (1|Population/ReplicateCage), data = filter(d_LS_P_surv, Supervisor.PI == "Flatt", Sex == "M"))
capture.output(summary(LS_P_M_coxme_Flatt), file = "Starvation/LS_P_M_coxme_Flatt_sum.txt")
capture.output(anova(LS_P_M_coxme_Flatt), file = "Starvation/LS_P_M_coxme_Flatt")

LS_L_F_coxme_Parsch <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_LS_L_surv, Supervisor.PI == "Parsch", Sex == "F"))
capture.output(summary(LS_L_F_coxme_Parsch), file = "Starvation/LS_L_F_coxme_Parsch_sum.txt")
capture.output(anova(LS_L_F_coxme_Parsch), file = "Starvation/LS_L_F_coxme_Parsch.txt")

LS_L_M_coxme_Parsch <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_LS_L_surv, Supervisor.PI == "Parsch", Sex == "M"))
capture.output(summary(LS_L_M_coxme_Parsch), file = "Starvation/LS_L_M_coxme_Parsch_sum.txt")
capture.output(anova(LS_L_M_coxme_Parsch), file = "Starvation/LS_L_M_coxme_Parsch.txt")

LS_L_F_coxme_Pasyukova <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_LS_L_surv, Supervisor.PI == "Pasyukova", Sex == "F"))
capture.output(summary(LS_L_F_coxme_Pasyukova), file = "Starvation/LS_L_F_coxme_Pasyukova_sum.txt")
capture.output(anova(LS_L_F_coxme_Pasyukova), file = "Starvation/LS_L_F_coxme_Pasyukova.txt")

LS_L_M_coxme_Pasyukova <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_LS_L_surv, Supervisor.PI == "Pasyukova", Sex == "M"))
capture.output(summary(LS_L_M_coxme_Pasyukova), file = "Starvation/LS_L_M_coxme_Pasyukova_sum.txt")
capture.output(anova(LS_L_M_coxme_Pasyukova), file = "Starvation/LS_L_M_coxme_Pasyukova.txt")

Lab correlations

intersect_Line_LS <- intersect(d_LS_L$Line[d_LS_L$Supervisor.PI == "Parsch"], 
                           d_LS_L$Line[d_LS_L$Supervisor.PI == "Pasyukova"])

write(intersect_Line_LS, file ="Lifespan/intersect_Line_LS", sep = "\t")

intersect_Pop_LS <- Reduce(intersect, list(d_LS_M$Population[d_LS_M$Supervisor.PI == "Parsch"],
                              d_LS_L$Line[d_LS_L$Supervisor.PI == "Flatt"],
                           d_LS_M$Population[d_LS_M$Supervisor.PI == "Pasyukova"]))

write(intersect_Pop_LS, file ="Lifespan/intersect_Pop_LS", sep = "\t")

There are 0 populations and 133 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

LS_fitted <- combineFitted(labs = rep(c("Flatt", "Parsch", "Pasyukova"), 2), sex = rep(c("F", "M"), each = 3), models = list(LS_P_F_Flatt_lmer, LS_L_F_Parsch_lmer, LS_L_F_Pasyukova_lmer, LS_P_M_Flatt_lmer, LS_L_M_Parsch_lmer, LS_L_M_Pasyukova_lmer))

p_LS_F_lab_cor <- function() scatterPlotMatrix(LS_fitted, sex = "F")
p_LS_M_lab_cor <- function() scatterPlotMatrix(LS_fitted, sex = "M")

p_LS_F_lab_cor()

pdf(file="Lifespan/p_LS_F_lab_cor.pdf",width=8, height=8)
p_LS_F_lab_cor()
invisible(dev.off())

p_LS_M_lab_cor()

pdf(file="Lifespan/p_LS_M_lab_cor.pdf",width=8, height=8)
p_LS_M_lab_cor()
invisible(dev.off())

Cold-shock mortality

dir.create(file.path("ColdShock"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Gonzalez Lab : Llewellyn Green, Josefa Gonzalez, Miriam Merenciano

Kozeretska Lab : Svitlana Serga, Alexandra Protsenko, Oleksandr Maistrenko, Iryna Kozeretska

Vieira Lab : Jorge Vieira, Cristina P. Vieira, Pedro Duque, Tânia Dias

Data Reformatting

d_CSM <- read.csv("MasterSheets_Oct21_git/CSM_MasterSheet_Oct21.csv")

str(d_CSM)

## 'data.frame':    2584 obs. of  16 variables:
##  $ Supervisor.PI   : chr  "Gonzalez" "Gonzalez" "Gonzalez" "Gonzalez" ...
##  $ Diet            : chr  "S" "S" "S" "S" ...
##  $ Batch           : int  1 2 3 1 3 4 1 2 3 1 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : chr  "AK1" "AK1" "AK1" "AK10" ...
##  $ Sex             : chr  "M" "M" "M" "M" ...
##  $ ReplicateVialOld: int  1 2 3 1 2 3 1 2 3 1 ...
##  $ ReplicateVial   : chr  "Gonzalez_1_AK1_M_1" "Gonzalez_2_AK1_M_2" "Gonzalez_3_AK1_M_3" "Gonzalez_1_AK10_M_1" ...
##  $ Total           : int  20 19 18 20 20 20 16 20 18 16 ...
##  $ Dead            : int  14 11 3 16 4 11 8 17 4 16 ...
##  $ CSM_PropDead    : num  70 57.9 16.7 80 20 55 50 85 22.2 100 ...
##  $ CSM_PropDead_ED : num  0.7 0.579 0.167 0.8 0.2 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...

d_CSM$Supervisor.PI <- as.factor(d_CSM$Supervisor.PI)
d_CSM$Diet <- as.factor(d_CSM$Diet)
d_CSM$Batch <- as.factor(d_CSM$Batch)
d_CSM$Population_Lat <- factor(d_CSM$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_CSM$Population_Lon <- factor(d_CSM$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_CSM$Population_Alt <- factor(d_CSM$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_CSM$Line <- as.factor(d_CSM$Line)
d_CSM$Sex <- as.factor(d_CSM$Sex)
d_CSM$ReplicateVial <- as.factor(d_CSM$ReplicateVial)
d_CSM$CSM_PropDead_ED <- as.numeric(d_CSM$CSM_PropDead_ED)

str(d_CSM)

## 'data.frame':    2584 obs. of  19 variables:
##  $ Supervisor.PI   : Factor w/ 3 levels "Gonzalez","Kozeretska",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet            : Factor w/ 2 levels "NS","S": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Batch           : Factor w/ 32 levels "1","2","3","4",..: 1 2 3 1 3 4 1 2 3 1 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : Factor w/ 172 levels "AK1","AK10","AK11",..: 1 1 1 2 2 2 3 3 3 4 ...
##  $ Sex             : Factor w/ 2 levels "F","M": 2 2 2 2 2 2 2 2 2 2 ...
##  $ ReplicateVialOld: int  1 2 3 1 2 3 1 2 3 1 ...
##  $ ReplicateVial   : Factor w/ 2584 levels "Gonzalez_1_AK1_F_1",..: 2 152 342 4 344 584 6 154 346 8 ...
##  $ Total           : int  20 19 18 20 20 20 16 20 18 16 ...
##  $ Dead            : int  14 11 3 16 4 11 8 17 4 16 ...
##  $ CSM_PropDead    : num  70 57.9 16.7 80 20 55 50 85 22.2 100 ...
##  $ CSM_PropDead_ED : num  0.7 0.579 0.167 0.8 0.2 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat  : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon  : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt  : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_CSM_F <-subset(d_CSM,Sex=='F')
d_CSM_M <-subset(d_CSM,Sex=='M')

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

table_CSM_Line_wbatch <- write.csv(d_CSM %>% group_by(Supervisor.PI, Batch, Sex,Population, Line) %>% 
                                   summarise_at(vars(CSM_PropDead_ED), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "ColdShock/table_CSM_Line_wbatch.csv", row.names = T)
table_CSM_Line_wbatch <- read.csv("ColdShock/table_CSM_Line_wbatch.csv")

Descriptive statistics at the line level, without batch information :

table_CSM_Line_wobatch <- write.csv(d_CSM %>% group_by(Supervisor.PI, Sex,Population, Line) %>% 
                                   summarise_at(vars(CSM_PropDead_ED), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "ColdShock/table_CSM_Line_wobatch.csv", row.names = T)
table_CSM_Line_wobatch <- read.csv("ColdShock/table_CSM_Line_wobatch.csv")

Descriptive statistics at the population level, with batch information :

table_CSM_Pop_wbatch <- write.csv(d_CSM %>% group_by(Supervisor.PI, Batch, Sex, Population) %>% 
                                  summarise_at(vars(CSM_PropDead_ED), 
                                               list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                    SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                          file = "ColdShock/table_CSM_Pop_wbatch.csv", row.names = T)

table_CSM_Pop_wbatch <- read.csv("ColdShock/table_CSM_Pop_wbatch.csv")

Descriptive statistics at the population level, without batch information :

table_CSM_Pop_wobatch <- write.csv(d_CSM %>% group_by(Supervisor.PI, Sex, Population) %>% 
                                  summarise_at(vars(CSM_PropDead_ED), 
                                               list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                    SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                          file = "ColdShock/table_CSM_Pop_wobatch.csv", row.names = T)

table_CSM_Pop_wobatch <- read.csv("ColdShock/table_CSM_Pop_wobatch.csv")

Plots and Linear Models by Lab

Overview

min_CSM <- min(d_CSM$CSM_PropDead_ED)
max_CSM <- max(d_CSM$CSM_PropDead_ED)

y-axis is scaled by the minimum (0) and maximum (1) values in the full data set.

p_CSM_F <- ggplot(data = d_CSM_F, aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_F", x="Population", y = "Prop_dead") + ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_F

pdf(file="ColdShock/p_CSM_F.pdf",width=8, height=5)
p_CSM_F
dev.off()

## quartz_off_screen 
##                 2

p_CSM_M <- ggplot(data = d_CSM_M, aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_M", x="Population", y = "Prop_dead") +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_M

pdf(file="ColdShock/p_CSM_M.pdf",width=8, height=5)
p_CSM_M
dev.off()

## quartz_off_screen 
##                 2

Gonzalez Lab

CSM_F_lmer_Gonzalez <- lmer(CSM_PropDead_ED ~ -1 + Population  + (1| Line : Population) +
                           (1|Batch), data = d_CSM_F[d_CSM_F$Supervisor.PI == "Gonzalez",])
capture.output(summary(CSM_F_lmer_Gonzalez),file = "ColdShock/CSM_F_lmer_sum_Gonzalez.txt")
capture.output(anova(CSM_F_lmer_Gonzalez),file = "ColdShock/CSM_F_lmer_Gonzalez.txt")
capture.output(emmeans(CSM_F_lmer_Gonzalez, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ColdShock/CSM_F_lmer_Gonzalez_tk.txt")

anova(CSM_F_lmer_Gonzalez)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 17.836  1.9818     9 52.799  139.46 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(CSM_F_lmer_Gonzalez)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CSM_PropDead_ED ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: d_CSM_F[d_CSM_F$Supervisor.PI == "Gonzalez", ]
## 
## REML criterion at convergence: -520.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.5262 -0.3749  0.0953  0.5610  2.1847 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.009816 0.09908 
##  Batch           (Intercept) 0.004154 0.06446 
##  Residual                    0.014210 0.11921 
## Number of obs: 544, groups:  Line:Population, 160; Batch, 9
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.90153    0.03466 38.43295   26.01   <2e-16 ***
## PopulationGI  0.77871    0.03837 53.62814   20.29   <2e-16 ***
## PopulationKA  0.85685    0.03771 50.33708   22.72   <2e-16 ***
## PopulationMA  0.85223    0.03478 39.43863   24.50   <2e-16 ***
## PopulationMU  0.88218    0.03498 40.02650   25.22   <2e-16 ***
## PopulationRE  0.84133    0.03679 47.51782   22.87   <2e-16 ***
## PopulationUM  0.77992    0.03896 56.11118   20.02   <2e-16 ***
## PopulationVA  0.83379    0.03497 39.55000   23.85   <2e-16 ***
## PopulationYE  0.87437    0.03441 37.97353   25.41   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.380                                                  
## PopulatinKA 0.382  0.347                                           
## PopulatinMA 0.410  0.373  0.376                                    
## PopulatinMU 0.414  0.375  0.378  0.406                             
## PopulatinRE 0.378  0.343  0.349  0.376  0.375                      
## PopulatinUM 0.370  0.336  0.338  0.364  0.366  0.336               
## PopulatinVA 0.420  0.380  0.381  0.409  0.414  0.376  0.369        
## PopulatinYE 0.408  0.371  0.377  0.406  0.405  0.380  0.364  0.406

CSM_M_lmer_Gonzalez <- lmer(CSM_PropDead_ED ~ -1 + Population  + (1| Line : Population) +
                                    (1|Batch), data = d_CSM_M[d_CSM_M$Supervisor.PI == "Gonzalez",])
capture.output(summary(CSM_M_lmer_Gonzalez),file = "ColdShock/CSM_M_lmer_sum_Gonzalez.txt")
capture.output(anova(CSM_M_lmer_Gonzalez),file = "ColdShock/CSM_M_lmer_Gonzalez.txt")
capture.output(emmeans(CSM_M_lmer_Gonzalez, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ColdShock/CSM_M_lmer_Gonzalez_tk.txt")

anova(CSM_M_lmer_Gonzalez)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 4.0718 0.45242     9 46.507  14.742 6.315e-11 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(CSM_M_lmer_Gonzalez)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CSM_PropDead_ED ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: d_CSM_M[d_CSM_M$Supervisor.PI == "Gonzalez", ]
## 
## REML criterion at convergence: -119.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.2121 -0.6190  0.0413  0.5522  2.3695 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.01754  0.1324  
##  Batch           (Intercept) 0.02650  0.1628  
##  Residual                    0.03069  0.1752  
## Number of obs: 544, groups:  Line:Population, 160; Batch, 9
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.58500    0.06612 16.18080   8.848 1.34e-07 ***
## PopulationGI  0.63730    0.07001 20.14264   9.103 1.41e-08 ***
## PopulationKA  0.66190    0.06920 19.25441   9.565 9.48e-09 ***
## PopulationMA  0.61389    0.06626 16.34959   9.265 6.54e-08 ***
## PopulationMU  0.59773    0.06647 16.54743   8.992 8.96e-08 ***
## PopulationRE  0.59781    0.06817 18.25751   8.769 5.77e-08 ***
## PopulationUM  0.53765    0.07054 20.73466   7.622 1.93e-07 ***
## PopulationVA  0.56016    0.06648 16.52017   8.426 2.21e-07 ***
## PopulationYE  0.71131    0.06582 15.93957  10.806 9.61e-09 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.660                                                  
## PopulatinKA 0.664  0.629                                           
## PopulatinMA 0.691  0.655  0.660                                    
## PopulatinMU 0.693  0.655  0.660  0.687                             
## PopulatinRE 0.665  0.630  0.637  0.663  0.662                      
## PopulatinUM 0.652  0.617  0.622  0.647  0.648  0.623               
## PopulatinVA 0.696  0.658  0.662  0.689  0.691  0.662  0.650        
## PopulatinYE 0.692  0.655  0.663  0.690  0.688  0.668  0.649  0.689

p_CSM_F_Gonzalez <- ggplot(data = (subset(d_CSM_F, Supervisor.PI == "Gonzalez")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_F_Gonzalez", x="Population", y = "CSM (PropSur)") + ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_F_Gonzalez

pdf(file="ColdShock/p_CSM_F_Gonzalez.pdf",width=8, height=5)
p_CSM_F_Gonzalez
dev.off()

## quartz_off_screen 
##                 2

p_CSM_F_pop_Gonzalez <- ggplot(data = (subset(d_CSM_F, Supervisor.PI == "Gonzalez")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_F_pop_Gonzalez", x="Population", y = "CSM (PropSur)") +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_F_pop_Gonzalez

pdf(file="ColdShock/p_CSM_F_pop_Gonzalez.pdf",width=8, height=5)
p_CSM_F_pop_Gonzalez
dev.off()

## quartz_off_screen 
##                 2

p_CSM_M_Gonzalez <- ggplot(data = (subset(d_CSM_M, Supervisor.PI == "Gonzalez")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_M_Gonzalez", x="Population", y = "CSM (PropSur)") + ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_M_Gonzalez

pdf(file="ColdShock/p_CSM_M_Gonzalez.pdf",width=8, height=5)
p_CSM_M_Gonzalez
dev.off()

## quartz_off_screen 
##                 2

p_CSM_M_pop_Gonzalez <- ggplot(data = (subset(d_CSM_M, Supervisor.PI == "Gonzalez")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_M_pop_Gonzalez", x="Population", y = "CSM (PropSur)") + ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_M_pop_Gonzalez

pdf(file="ColdShock/p_CSM_M_pop_Gonzalez.pdf",width=8, height=5)
p_CSM_M_pop_Gonzalez
dev.off()

## quartz_off_screen 
##                 2

Kozeretska Lab

CSM_F_lmer_Kozeretska <- lmer(CSM_PropDead_ED ~ -1 + Population  + (1| Line : Population) +
                                      (1|Batch), data = d_CSM_F[d_CSM_F$Supervisor.PI == "Kozeretska",])
capture.output(summary(CSM_F_lmer_Kozeretska),file = "ColdShock/CSM_F_lmer_sum_Kozeretska.txt")
capture.output(anova(CSM_F_lmer_Kozeretska),file = "ColdShock/CSM_F_lmer_Kozeretska.txt")
capture.output(emmeans(CSM_F_lmer_Kozeretska, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ColdShock/CSM_F_lmer_Kozeretska_tk.txt")

anova(CSM_F_lmer_Kozeretska)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value Pr(>F)
## Population 2.2818 0.25353     9     2    3.26 0.2568

summary(CSM_F_lmer_Kozeretska)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CSM_PropDead_ED ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: d_CSM_F[d_CSM_F$Supervisor.PI == "Kozeretska", ]
## 
## REML criterion at convergence: 134
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -1.6679 -0.7659 -0.1923  0.7403  2.0452 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.01410  0.1188  
##  Batch           (Intercept) 0.01886  0.1373  
##  Residual                    0.07777  0.2789  
## Number of obs: 244, groups:  Line:Population, 130; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)  
## PopulationAK   0.5828     0.1126 1.5833   5.174   0.0563 .
## PopulationGI   0.5178     0.1265 2.5161   4.093   0.0366 *
## PopulationKA   0.4528     0.1183 1.9321   3.826   0.0655 .
## PopulationMA   0.4791     0.1148 1.7063   4.173   0.0690 .
## PopulationMU   0.5249     0.1140 1.6577   4.605   0.0620 .
## PopulationRE   0.4998     0.1182 1.9225   4.228   0.0553 .
## PopulationUM   0.4336     0.1206 2.0744   3.595   0.0657 .
## PopulationVA   0.5057     0.1122 1.5551   4.506   0.0715 .
## PopulationYE   0.5176     0.1180 1.9080   4.385   0.0525 .
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.670                                                  
## PopulatinKA 0.708  0.630                                           
## PopulatinMA 0.746  0.659  0.695                                    
## PopulatinMU 0.752  0.664  0.700  0.742                             
## PopulatinRE 0.716  0.635  0.674  0.704  0.709                      
## PopulatinUM 0.710  0.627  0.662  0.700  0.705  0.670               
## PopulatinVA 0.768  0.677  0.711  0.758  0.764  0.723  0.721        
## PopulatinYE 0.723  0.640  0.676  0.713  0.718  0.683  0.678  0.734

CSM_M_lmer_Kozeretska <- lmer(CSM_PropDead_ED ~ -1 + Population  + (1| Line : Population) +
                                      (1|Batch), data = d_CSM_M[d_CSM_M$Supervisor.PI == "Kozeretska",])
capture.output(summary(CSM_M_lmer_Kozeretska),file = "ColdShock/CSM_M_lmer_sum_Kozeretska.txt")
capture.output(anova(CSM_M_lmer_Kozeretska),file = "ColdShock/CSM_M_lmer_Kozeretska.txt")
capture.output(emmeans(CSM_M_lmer_Kozeretska, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ColdShock/CSM_M_tk_Kozeretska_tk.txt")

anova(CSM_M_lmer_Kozeretska)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value Pr(>F)
## Population 4.7634 0.52927     9     2  6.8565 0.1337

summary(CSM_M_lmer_Kozeretska)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CSM_PropDead_ED ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: d_CSM_M[d_CSM_M$Supervisor.PI == "Kozeretska", ]
## 
## REML criterion at convergence: 108.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.3573 -0.7476  0.1784  0.7742  1.5219 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.003936 0.06274 
##  Batch           (Intercept) 0.017486 0.13223 
##  Residual                    0.077193 0.27784 
## Number of obs: 244, groups:  Line:Population, 130; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)  
## PopulationAK   0.7641     0.1067 1.5015   7.162   0.0387 *
## PopulationGI   0.7930     0.1187 2.3020   6.679   0.0149 *
## PopulationKA   0.6545     0.1115 1.7946   5.869   0.0356 *
## PopulationMA   0.7342     0.1086 1.6100   6.761   0.0360 *
## PopulationMU   0.6656     0.1079 1.5662   6.172   0.0440 *
## PopulationRE   0.6864     0.1112 1.7735   6.173   0.0335 *
## PopulationUM   0.6538     0.1133 1.9053   5.772   0.0321 *
## PopulationVA   0.6958     0.1063 1.4737   6.548   0.0459 *
## PopulationYE   0.7831     0.1115 1.7895   7.024   0.0262 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.698                                                  
## PopulatinKA 0.736  0.661                                           
## PopulatinMA 0.770  0.688  0.723                                    
## PopulatinMU 0.776  0.693  0.728  0.766                             
## PopulatinRE 0.744  0.667  0.706  0.733  0.738                      
## PopulatinUM 0.738  0.659  0.693  0.728  0.734  0.702               
## PopulatinVA 0.791  0.706  0.739  0.782  0.788  0.752  0.749        
## PopulatinYE 0.748  0.669  0.704  0.738  0.744  0.713  0.707  0.759

p_CSM_F_Kozeretska <- ggplot(data = (subset(d_CSM_F, Supervisor.PI == "Kozeretska")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_F_Kozeretska", x="Population", y = "CSM (PropSur)") + ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_F_Kozeretska

pdf(file="ColdShock/p_CSM_F_Kozeretska.pdf",width=8, height=5)
p_CSM_F_Kozeretska
dev.off()

## quartz_off_screen 
##                 2

p_CSM_F_pop_Kozeretska <- ggplot(data = (subset(d_CSM_F, Supervisor.PI == "Kozeretska")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_F_pop_Kozeretska", x="Population", y = "CSM (PropSur)") +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_F_pop_Kozeretska

pdf(file="ColdShock/p_CSM_F_pop_Kozeretska.pdf",width=8, height=5)
p_CSM_F_pop_Kozeretska
dev.off()

## quartz_off_screen 
##                 2

p_CSM_M_Kozeretska <- ggplot(data = (subset(d_CSM_M, Supervisor.PI == "Kozeretska")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_M_Kozeretska", x="Population", y = "CSM (PropSur)")  +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_M_Kozeretska

pdf(file="ColdShock/p_CSM_M_Kozeretska.pdf",width=8, height=5)
p_CSM_M_Kozeretska
dev.off()

## quartz_off_screen 
##                 2

p_CSM_M_pop_Kozeretska <- ggplot(data = (subset(d_CSM_M, Supervisor.PI == "Kozeretska")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_M_pop_Kozeretska", x="Population", y = "CSM (PropSur)")  +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_M_pop_Kozeretska

pdf(file="ColdShock/p_CSM_M_pop_Kozeretska.pdf",width=8, height=5)
p_CSM_M_pop_Kozeretska
dev.off()

## quartz_off_screen 
##                 2

Vieira Lab

CSM_F_lmer_Vieira <- lmer(CSM_PropDead_ED ~ -1 + Population  + (1| Line : Population) +
                                  (1|Batch), data = d_CSM_F[d_CSM_F$Supervisor.PI == "Vieira",])
capture.output(summary(CSM_F_lmer_Vieira),file = "ColdShock/CSM_F_lmer_sum_Vieira.txt")
capture.output(anova(CSM_F_lmer_Vieira),file = "ColdShock/CSM_F_lmer_Vieira.txt")
capture.output(emmeans(CSM_F_lmer_Vieira, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ColdShock/CSM_F_lmer_Vieira_tk.txt")

anova(CSM_F_lmer_Vieira)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 69.384  7.7093     9 24.538  660.91 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(CSM_F_lmer_Vieira)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CSM_PropDead_ED ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: d_CSM_F[d_CSM_F$Supervisor.PI == "Vieira", ]
## 
## REML criterion at convergence: -527.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.4994 -0.3263  0.1582  0.4377  2.3707 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev.
##  Line:Population (Intercept) 0.0125606 0.11207 
##  Batch           (Intercept) 0.0008187 0.02861 
##  Residual                    0.0116647 0.10800 
## Number of obs: 504, groups:  Line:Population, 168; Batch, 32
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.85973    0.03082 43.39219   27.90  < 2e-16 ***
## PopulationGI  0.84992    0.03775 60.10192   22.51  < 2e-16 ***
## PopulationKA  0.89367    0.03131 30.10543   28.54  < 2e-16 ***
## PopulationMA  0.86227    0.03164 22.95302   27.25  < 2e-16 ***
## PopulationMU  0.87931    0.03168 25.75001   27.76  < 2e-16 ***
## PopulationRE  0.86198    0.03582 18.21320   24.07 2.93e-15 ***
## PopulationUM  0.78922    0.03173 40.56536   24.87  < 2e-16 ***
## PopulationVA  0.93223    0.03202 19.67021   29.11  < 2e-16 ***
## PopulationYE  0.87808    0.03206 18.99480   27.39  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.037                                                  
## PopulatinKA 0.001  0.001                                           
## PopulatinMA 0.000  0.000  0.007                                    
## PopulatinMU 0.068  0.035  0.001  0.000                             
## PopulatinRE 0.000  0.000  0.011  0.000  0.000                      
## PopulatinUM 0.015  0.017  0.034  0.000  0.038  0.000               
## PopulatinVA 0.000  0.000  0.000  0.001  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.010  0.000  0.000  0.000  0.076

CSM_M_lmer_Vieira <- lmer(CSM_PropDead_ED ~ -1 + Population  + (1| Line : Population) +
                                  (1|Batch), data = d_CSM_M[d_CSM_M$Supervisor.PI == "Vieira",])
capture.output(summary(CSM_M_lmer_Vieira),file = "ColdShock/CSM_M_lmer_sum_Vieira.txt")
capture.output(anova(CSM_M_lmer_Vieira),file = "ColdShock/CSM_M_lmer_Vieira.txt")
capture.output(emmeans(CSM_M_lmer_Vieira, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ColdShock/CSM_M_lmer_Vieira_tk.txt")

anova(CSM_M_lmer_Vieira)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 28.078  3.1198     9 44.934  169.51 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(CSM_M_lmer_Vieira)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CSM_PropDead_ED ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: d_CSM_M[d_CSM_M$Supervisor.PI == "Vieira", ]
## 
## REML criterion at convergence: -249.1
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.8114 -0.3899  0.1187  0.4494  2.5947 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.027998 0.16733 
##  Batch           (Intercept) 0.005384 0.07337 
##  Residual                    0.018405 0.13566 
## Number of obs: 504, groups:  Line:Population, 168; Batch, 32
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.71755    0.04969 62.48430   14.44  < 2e-16 ***
## PopulationGI  0.72299    0.05985 79.14734   12.08  < 2e-16 ***
## PopulationKA  0.74224    0.05144 50.71725   14.43  < 2e-16 ***
## PopulationMA  0.71924    0.05311 37.18225   13.54 5.98e-16 ***
## PopulationMU  0.87253    0.05217 53.61528   16.73  < 2e-16 ***
## PopulationRE  0.76734    0.06113 30.70270   12.55 1.26e-13 ***
## PopulationUM  0.67972    0.05123 63.70981   13.27  < 2e-16 ***
## PopulationVA  0.78489    0.05407 36.19392   14.52  < 2e-16 ***
## PopulationYE  0.89860    0.05421 36.47581   16.58  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.103                                                  
## PopulatinKA 0.004  0.004                                           
## PopulatinMA 0.000  0.000  0.023                                    
## PopulatinMU 0.171  0.094  0.008  0.000                             
## PopulatinRE 0.000  0.000  0.032  0.001  0.000                      
## PopulatinUM 0.047  0.045  0.086  0.002  0.095  0.003               
## PopulatinVA 0.000  0.000  0.000  0.006  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.001  0.030  0.000  0.000  0.000  0.191

p_CSM_F_Vieira <- ggplot(data = (subset(d_CSM_F, Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_F_Vieira", x="Population", y = "CSM (PropSur)") +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_F_Vieira

pdf(file="ColdShock/p_CSM_F_Vieira.pdf",width=8, height=5)
p_CSM_F_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_CSM_F_pop_Vieira <- ggplot(data = (subset(d_CSM_F, Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_F_pop_Vieira", x="Population", y = "CSM (PropSur)") +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_F_pop_Vieira

pdf(file="ColdShock/p_CSM_F_pop_Vieira.pdf",width=8, height=5)
p_CSM_F_pop_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_CSM_M_Vieira <- ggplot(data = (subset(d_CSM_M, Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_M_Vieira", x="Population", y = "CSM (PropSur)") +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_M_Vieira

pdf(file="ColdShock/p_CSM_M_Vieira.pdf",width=8, height=5)
p_CSM_M_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_CSM_M_pop_Vieira <- ggplot(data = (subset(d_CSM_M, Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=CSM_PropDead_ED, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CSM_M_pop_Vieira", x="Population", y = "CSM (PropSur)") +  ylim(c(min_CSM, max_CSM))+ droseu_theme

p_CSM_M_pop_Vieira

pdf(file="ColdShock/p_CSM_M_pop_Vieira.pdf",width=8, height=5)
p_CSM_M_pop_Vieira
dev.off()

## quartz_off_screen 
##                 2

Lab correlations

intersect_Line_CSM  <- Reduce(intersect, list(d_CSM$Line[d_CSM$Supervisor.PI == "Gonzalez"],
                                         d_CSM$Line[d_CSM$Supervisor.PI == "Vieira"],
                                         d_CSM$Line[d_CSM$Supervisor.PI == "Kozeretska"]))
write(intersect_Line_CSM, file ="ColdShock/intersect_Line_CSM.txt", sep = "\t")

intersect_Pop_CSM  <- Reduce(intersect, list(d_CSM$Population[d_CSM$Supervisor.PI == "Gonzalez"],
                                         d_CSM$Population[d_CSM$Supervisor.PI == "Vieira"],
                                         d_CSM$Population[d_CSM$Supervisor.PI == "Kozeretska"]))
write(intersect_Pop_CSM, file ="ColdShock/intersect_Pop_CSM.txt", sep = "\t")

There are 9 populations and 126 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

CSM_fitted <- combineFitted(labs = rep(c("Gonzalez", "Kozeretska", "Vieira"), 2), sex = rep(c("F", "M"), each = 3), models = list(CSM_F_lmer_Gonzalez, CSM_F_lmer_Kozeretska, CSM_F_lmer_Vieira, CSM_M_lmer_Gonzalez, CSM_M_lmer_Kozeretska, CSM_M_lmer_Vieira))


p_CSM_F_lab_cor <- function() scatterPlotMatrix(CSM_fitted, sex = "F")
p_CSM_M_lab_cor <- function() scatterPlotMatrix(CSM_fitted, sex = "M")

p_CSM_F_lab_cor()

pdf(file="ColdShock/p_CSM_F_lab_cor.pdf", width=8, height=8)
p_CSM_F_lab_cor()
invisible(dev.off())

Chill-coma recovery time

dir.create(file.path("ChillComa"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Mensh Lab : Florencia Putero, Lucas Kreiman, Julian Mensh

Vieira Lab : Jorge Vieira, Cristina P. Vieira, Pedro Duque, Tânia Dias

Data Reformatting

d_CCRT <- read.csv("MasterSheets_Oct21_git/CCRT_MasterSheet_Oct21.csv")

str(d_CCRT)

## 'data.frame':    2016 obs. of  15 variables:
##  $ Supervisor.PI   : chr  "Vieira" "Vieira" "Vieira" "Vieira" ...
##  $ Diet            : chr  "S" "S" "S" "S" ...
##  $ Batch           : int  4 4 4 4 4 4 4 4 4 4 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Sex             : chr  "F" "F" "F" "F" ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : chr  "Vieira_4_AK1_F_1" "Vieira_4_AK1_F_1" "Vieira_4_AK1_F_1" "Vieira_4_AK1_F_1" ...
##  $ Individual      : int  1 2 3 4 5 6 1 2 3 4 ...
##  $ CCRT_seconds    : int  2381 1902 1847 1640 2202 1444 1550 1900 1505 1524 ...
##  $ Censor          : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...

d_CCRT$Supervisor.PI <- as.factor(d_CCRT$Supervisor.PI)
d_CCRT$Diet <- as.factor(d_CCRT$Diet)
d_CCRT$Batch <- as.factor(d_CCRT$Batch)
d_CCRT$Population_Lat <- factor(d_CCRT$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_CCRT$Population_Lon <- factor(d_CCRT$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_CCRT$Population_Alt <- factor(d_CCRT$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_CCRT$Line <- as.factor(d_CCRT$Line)
d_CCRT$Sex <- as.factor(d_CCRT$Sex)
d_CCRT$ReplicateVial <- as.factor(d_CCRT$ReplicateVial)
d_CCRT$CCRT_seconds <- as.numeric(d_CCRT$CCRT_seconds)
d_CCRT$Censor <- as.numeric(d_CCRT$Censor)

str(d_CCRT)

## 'data.frame':    2016 obs. of  18 variables:
##  $ Supervisor.PI   : Factor w/ 1 level "Vieira": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet            : Factor w/ 1 level "S": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch           : Factor w/ 27 levels "1","2","3","4",..: 4 4 4 4 4 4 4 4 4 4 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : Factor w/ 168 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex             : Factor w/ 2 levels "F","M": 1 1 1 1 1 1 2 2 2 2 ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : Factor w/ 336 levels "Vieira_1_UM13_F_1",..: 235 235 235 235 235 235 236 236 236 236 ...
##  $ Individual      : int  1 2 3 4 5 6 1 2 3 4 ...
##  $ CCRT_seconds    : num  2381 1902 1847 1640 2202 ...
##  $ Censor          : num  0 0 0 0 0 0 0 0 0 0 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat  : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon  : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt  : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_CCRT_F <-subset(d_CCRT,Sex=='F')
d_CCRT_M <-subset(d_CCRT,Sex=='M')

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

table_CCRT_Line_wbatch <- write.csv(d_CCRT %>% group_by(Supervisor.PI, Batch, Sex,Population, Line) %>% 
                                    summarise_at(vars(CCRT_seconds), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "ChillComa/table_CCRT_Line_wbatch.csv", row.names = T)

table_CCRT_Line_wbatch <- read.csv("ChillComa/table_CCRT_Line_wbatch.csv")

Descriptive statistics at the line level, without batch information :

table_CCRT_Line_wobatch <- write.csv(d_CCRT %>% group_by(Supervisor.PI, Sex,Population, Line) %>% 
                                    summarise_at(vars(CCRT_seconds), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "ChillComa/table_CCRT_Line_wobatch.csv", row.names = T)

table_CCRT_Line_wobatch <- read.csv("ChillComa/table_CCRT_Line_wobatch.csv")

Descriptive statistics at the population level, with batch information :

table_CCRT_pop_wbatch <- write.csv(d_CCRT %>% group_by(Supervisor.PI, Batch, Sex,Population) %>% 
                                    summarise_at(vars(CCRT_seconds), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "ChillComa/table_CCRT_pop_wbatch.csv", row.names = T)

table_CCRT_pop_wbatch <- read.csv("ChillComa/table_CCRT_pop_wbatch.csv")

Descriptive statistics at the population level, without batch information :

table_CCRT_pop_wobatch <- write.csv(d_CCRT %>% group_by(Supervisor.PI, Sex,Population) %>% 
                                    summarise_at(vars(CCRT_seconds), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "ChillComa/table_CCRT_pop_wobatch.csv", row.names = T)

table_CCRT_pop_wobatch <- read.csv("ChillComa/table_CCRT_pop_wobatch.csv")

Plots and Linear Models by Lab

Overview

min_CCRT <- min(d_CCRT$CCRT_seconds)
max_CCRT <- max(d_CCRT$CCRT_seconds)

y-axis is scaled by the minimum (0) and maximum (1) values in the full data set.

Vieira Lab

CCRT_F_lmer_Vieira <- lmer(CCRT_seconds ~ -1 + Population + (1|Population:Line) + (1|Batch) + (1|ReplicateVial:Line), data = filter(d_CCRT_F, Censor == "0" & Supervisor.PI == "Vieira"))
capture.output(summary(CCRT_F_lmer_Vieira),file = "ChillComa/CCRT_F_lmer_Vieira_sum.txt")
capture.output(anova(CCRT_F_lmer_Vieira),file = "ChillComa/CCRT_F_lmer_Vieira.txt")
capture.output(emmeans(CCRT_F_lmer_Vieira, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ChillComa/CCRT_F_lmer_Vieira_tk.txt")

model_means_CCRT_F_Vieira <- emmeans(object = CCRT_F_lmer_Vieira, specs = ~ Population)
model_means_CCRT_F_Vieira_cld <- cld(object = model_means_CCRT_F_Vieira, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(CCRT_F_lmer_Vieira)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 767340659 85260073     9 32.479  465.02 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(CCRT_F_lmer_Vieira)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CCRT_seconds ~ -1 + Population + (1 | Population:Line) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: filter(d_CCRT_F, Censor == "0" & Supervisor.PI == "Vieira")
## 
## REML criterion at convergence: 14874.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.1461 -0.5789 -0.1437  0.3652  5.0319 
## 
## Random effects:
##  Groups             Name        Variance Std.Dev.
##  ReplicateVial:Line (Intercept)    799.8  28.28  
##  Population:Line    (Intercept)  29469.4 171.67  
##  Batch              (Intercept)   3852.3  62.07  
##  Residual                       183347.9 428.19  
## Number of obs: 993, groups:  
## ReplicateVial:Line, 168; Population:Line, 168; Batch, 27
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  1582.02      62.96   38.07   25.13  < 2e-16 ***
## PopulationGI  1526.91      74.88   65.94   20.39  < 2e-16 ***
## PopulationKA  1388.14      62.08   42.64   22.36  < 2e-16 ***
## PopulationMA  1294.83      64.18   24.49   20.17  < 2e-16 ***
## PopulationMU  1532.41      66.30   17.90   23.11 8.89e-15 ***
## PopulationRE  1434.21      66.92   71.57   21.43  < 2e-16 ***
## PopulationUM  1714.00      64.27   37.74   26.67  < 2e-16 ***
## PopulationVA  1401.57      63.63   29.39   22.03  < 2e-16 ***
## PopulationYE  1482.00      61.88   37.25   23.95  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.150                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.018                                    
## PopulatinMU 0.006  0.040  0.001  0.000                             
## PopulatinRE 0.075  0.037  0.002  0.000  0.002                      
## PopulatinUM 0.005  0.009  0.036  0.001  0.017  0.051               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.143

# might need to remove batch or data -- singularity is not coming from reps 
CCRT_M_lmer_Vieira <- lmer(CCRT_seconds ~ -1 + Population + (1|Population:Line) + (1|Batch) + (1|ReplicateVial:Line), data = filter(d_CCRT_M, Censor == "0" & Supervisor.PI == "Vieira"))

## boundary (singular) fit: see ?isSingular

## Warning: Model failed to converge with 1 negative eigenvalue: -4.3e-05

capture.output(summary(CCRT_M_lmer_Vieira),file = "ChillComa/CCRT_M_lmer_Vieira_sum.txt")
capture.output(anova(CCRT_M_lmer_Vieira),file = "ChillComa/CCRT_M_lmer_Vieira.txt")
capture.output(emmeans(CCRT_M_lmer_Vieira, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "ChillComa/CCRT_M_lmer_Vieira_tk.txt")

model_means_CCRT_M_Vieira <- emmeans(object = CCRT_M_lmer_Vieira, specs = ~ Population)
model_means_CCRT_M_Vieira_cld <- cld(object = model_means_CCRT_M_Vieira, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

anova(CCRT_M_lmer_Vieira)

## Type III Analysis of Variance Table with Satterthwaite's method
##               Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 896220514 99580057     9 159.25  633.43 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(CCRT_M_lmer_Vieira)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CCRT_seconds ~ -1 + Population + (1 | Population:Line) + (1 |  
##     Batch) + (1 | ReplicateVial:Line)
##    Data: filter(d_CCRT_M, Censor == "0" & Supervisor.PI == "Vieira")
## 
## REML criterion at convergence: 14768.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.3458 -0.5562 -0.1358  0.3737  4.4516 
## 
## Random effects:
##  Groups             Name        Variance  Std.Dev. 
##  ReplicateVial:Line (Intercept) 1.539e+03 3.923e+01
##  Population:Line    (Intercept) 4.448e+04 2.109e+02
##  Batch              (Intercept) 8.019e-05 8.955e-03
##  Residual                       1.572e+05 3.965e+02
## Number of obs: 993, groups:  
## ReplicateVial:Line, 168; Population:Line, 168; Batch, 27
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  1563.51      60.09  157.52   26.02   <2e-16 ***
## PopulationGI  1844.86      74.54  157.52   24.75   <2e-16 ***
## PopulationKA  1445.65      60.50  161.63   23.89   <2e-16 ***
## PopulationMA  1329.62      60.19  158.53   22.09   <2e-16 ***
## PopulationMU  1694.41      60.09  157.52   28.20   <2e-16 ***
## PopulationRE  1563.65      67.33  158.78   23.22   <2e-16 ***
## PopulationUM  1578.36      61.87  159.65   25.51   <2e-16 ***
## PopulationVA  1505.40      60.30  159.54   24.97   <2e-16 ***
## PopulationYE  1653.14      60.61  162.70   27.28   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

p_CCRT_F_Vieira <- ggplot(data = filter(d_CCRT_F, (Censor == "0") & (Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=CCRT_seconds, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CCRT_F_Vieira", x="Population", y = "CCRT_seconds") + ylim(c(min_CCRT, max_CCRT))+ droseu_theme

p_CCRT_F_Vieira

pdf(file="ChillComa/p_CCRT_F_Vieira.pdf",width=8, height=5)
p_CCRT_F_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_CCRT_F_pop_Vieira <- ggplot(data = filter(d_CCRT_F, (Censor == "0") & (Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=CCRT_seconds, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CCRT_F_pop_Vieira", x="Population", y = "CCRT_seconds")+ ylim(c(min_CCRT, max_CCRT))+ geom_text(data = model_means_CCRT_F_Vieira_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_CCRT_F_pop_Vieira

pdf(file="ChillComa/p_CCRT_F_pop_Vieira.pdf",width=8, height=5)
p_CCRT_F_pop_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_CCRT_M_Vieira <- ggplot(data = filter(d_CCRT_M, (Censor == "0") & (Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=CCRT_seconds, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CCRT_M", x="Population", y = "CCRT_seconds") + ylim(c(min_CCRT, max_CCRT))+ droseu_theme

p_CCRT_M_Vieira

pdf(file="ChillComa/p_CCRT_M_Vieira.pdf",width=8, height=5)
p_CCRT_M_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_CCRT_M_pop_Vieira <- ggplot(data = filter(d_CCRT_M, (Censor == "0") & (Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=CCRT_seconds, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_CCRT_M_pop_Vieira", x="Population", y = "CCRT_seconds") + ylim(c(min_CCRT, max_CCRT))+ geom_text(data = model_means_CCRT_M_Vieira_cld, aes(x = Population, label = cld), y = 1, vjust = 1) + droseu_theme

p_CCRT_M_pop_Vieira

pdf(file="ChillComa/p_CCRT_M_pop_Vieira.pdf",width=8, height=5)
p_CCRT_M_pop_Vieira
dev.off()

## quartz_off_screen 
##                 2

Mensh Lab

Lab correlations

NA

Heat-shock mortality

dir.create(file.path("HeatShock"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Parsch Lab : Eliza Argyridou, Amanda Glaser-Schmitt, John Parsch

Vieira Lab : Jorge Vieira, Cristina P. Vieira, Pedro Duque, Tânia Dias

Data Reformatting

d_HSM <- read.csv("MasterSheets_Oct21_git/HSM_MasterSheet_Oct21.csv")

str(d_HSM)

## 'data.frame':    31244 obs. of  15 variables:
##  $ Supervisor.PI   : chr  "Vieira" "Vieira" "Vieira" "Vieira" ...
##  $ Diet            : chr  "S" "S" "S" "S" ...
##  $ Batch           : int  5 5 5 5 5 5 5 5 5 5 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Sex             : chr  "F" "F" "F" "F" ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : chr  "Vieira_5_AK1_F_1" "Vieira_5_AK1_F_1" "Vieira_5_AK1_F_1" "Vieira_5_AK1_F_1" ...
##  $ Individual      : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ TimeDeath_min   : int  240 240 300 300 300 330 360 360 360 390 ...
##  $ Censor          : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...

d_HSM$Supervisor.PI <- as.factor(d_HSM$Supervisor.PI)
d_HSM$Diet <- as.factor(d_HSM$Diet)
d_HSM$Batch <- as.factor(d_HSM$Batch)
d_HSM$Population_Lat <- factor(d_HSM$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_HSM$Population_Lon <- factor(d_HSM$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_HSM$Population_Alt <- factor(d_HSM$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_HSM$Line <- as.factor(d_HSM$Line)
d_HSM$Sex <- as.factor(d_HSM$Sex)
d_HSM$ReplicateVial <- as.factor(d_HSM$ReplicateVial)
d_HSM$TimeDeath_min <- as.numeric(d_HSM$TimeDeath_min)
d_HSM$Censor <- as.numeric(d_HSM$Censor)
str(d_HSM)

## 'data.frame':    31244 obs. of  18 variables:
##  $ Supervisor.PI   : Factor w/ 2 levels "Parsch","Vieira": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Diet            : Factor w/ 2 levels "NS","S": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Batch           : Factor w/ 32 levels "1","2","3","4",..: 5 5 5 5 5 5 5 5 5 5 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : Factor w/ 171 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex             : Factor w/ 2 levels "F","M": 1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : Factor w/ 1816 levels "Parsch_1_AK1_F_1",..: 1691 1691 1691 1691 1691 1691 1691 1691 1691 1691 ...
##  $ Individual      : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ TimeDeath_min   : num  240 240 300 300 300 330 360 360 360 390 ...
##  $ Censor          : num  0 0 0 0 0 0 0 0 0 0 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat  : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon  : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt  : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_HSM_F <-subset(d_HSM,Sex=='F')
d_HSM_M <-subset(d_HSM,Sex=='M')

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

table_HSM_Line_wbatch <- write.csv(d_HSM %>% group_by(Supervisor.PI, Batch, Sex,Population, Line) %>% 
                                    summarise_at(vars(TimeDeath_min), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "HeatShock/table_HSM_Line_wbatch.csv", row.names = T)
table_HSM_Line_wbatch <- read.csv("HeatShock/table_HSM_Line_wbatch.csv")

Descriptive statistics at the line level, without batch information :

table_HSM_Line_wobatch <- write.csv(d_HSM %>% group_by(Supervisor.PI, Sex,Population, Line) %>% 
                                    summarise_at(vars(TimeDeath_min), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "HeatShock/table_HSM_Line_wobatch.csv", row.names = T)
table_HSM_Line_wobatch <- read.csv("HeatShock/table_HSM_Line_wobatch.csv")

Descriptive statistics at the population level, with batch information :

table_HSM_Pop_wbatch <- write.csv(d_HSM %>% group_by(Supervisor.PI, Batch, Sex, Population) %>% 
                                   summarise_at(vars(TimeDeath_min), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "HeatShock/table_HSM_Pop_wbatch.csv", row.names = T)

table_HSM_Pop_wbatch <- read.csv("HeatShock/table_HSM_Pop_wbatch.csv")

Descriptive statistics at the population level, without batch information :

table_HSM_Pop_wobatch <- write.csv(d_HSM %>% group_by(Supervisor.PI, Sex, Population) %>% 
                                   summarise_at(vars(TimeDeath_min), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "HeatShock/table_HSM_Pop_wobatch.csv", row.names = T)
table_HSM_Pop_wobatch <- read.csv("HeatShock/table_HSM_Pop_wobatch.csv")

Plots and Linear Models by Lab

Overview

min_HSM <- min(d_HSM$TimeDeath_min)
max_HSM <- max(d_HSM$TimeDeath_min)

y-axis is scaled by the minimum (30) and maximum (505) values in the full data set.

p_HSM_F <- ggplot(data = filter(d_HSM_F, (Censor == "0")), aes(x=Population_Lat, y=TimeDeath_min, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_F", x="Population", y = "TimeDeath(min)")  + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_F

pdf(file="HeatShock/p_HSM_F.pdf",width=8, height=5)
p_HSM_F
dev.off()

## quartz_off_screen 
##                 2

p_HSM_M <- ggplot(data = filter(d_HSM_M, (Censor == "0")), aes(x=Population_Lat, y=TimeDeath_min, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_M", x="Population", y = "TimeDeath(min)")  + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_M

pdf(file="HeatShock/p_HSM_M.pdf",width=8, height=5)
p_HSM_M
dev.off()

## quartz_off_screen 
##                 2

Parsch Lab

HSM_F_lmer_Parsch <- lmer(TimeDeath_min ~ -1 + Population  + (1| Line : Population) + (1|Batch), data = filter(d_HSM_F, Censor == "0" & Supervisor.PI == "Parsch"))
capture.output(summary(HSM_F_lmer_Parsch),file = "HeatShock/HSM_F_lmer_sum_Parsch.txt")
capture.output(anova(HSM_F_lmer_Parsch),file = "HeatShock/HSM_F_lmer_Parsch.txt")
capture.output(emmeans(HSM_F_lmer_Parsch, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "HeatShock/HSM_F_lmer_Parsch_tk.txt")

anova(HSM_F_lmer_Parsch)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2247426  249714     9 43.663  41.562 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(HSM_F_lmer_Parsch)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TimeDeath_min ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: filter(d_HSM_F, Censor == "0" & Supervisor.PI == "Parsch")
## 
## REML criterion at convergence: 39308.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.9904 -0.5935  0.1317  0.6680  2.9181 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 2411     49.11   
##  Batch           (Intercept) 3662     60.51   
##  Residual                    6008     77.51   
## Number of obs: 3384, groups:  Line:Population, 133; Batch, 9
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   394.11      25.09  17.59   15.71 8.64e-12 ***
## PopulationGI   337.50      24.51  16.06   13.77 2.60e-10 ***
## PopulationKA   369.69      24.57  16.23   15.04 5.94e-11 ***
## PopulationMA   421.92      25.22  17.94   16.73 2.16e-12 ***
## PopulationMU   400.14      24.59  16.25   16.27 1.74e-11 ***
## PopulationRE   323.23      24.41  15.78   13.24 5.84e-10 ***
## PopulationUM   395.05      25.37  18.34   15.57 5.10e-12 ***
## PopulationVA   384.98      25.41  18.46   15.15 7.31e-12 ***
## PopulationYE   312.13      24.19  15.25   12.90 1.30e-09 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.673                                                  
## PopulatinKA 0.672  0.701                                           
## PopulatinMA 0.659  0.658  0.663                                    
## PopulatinMU 0.677  0.674  0.679  0.681                             
## PopulatinRE 0.651  0.666  0.666  0.650  0.666                      
## PopulatinUM 0.618  0.637  0.634  0.615  0.630  0.689               
## PopulatinVA 0.617  0.636  0.633  0.615  0.630  0.687  0.676        
## PopulatinYE 0.661  0.680  0.676  0.657  0.674  0.699  0.678  0.677

HSM_M_lmer_Parsch <- lmer(TimeDeath_min ~ -1 + Population  + (1| Line : Population) + (1|Batch), data = filter(d_HSM_M, Censor == "0" & Supervisor.PI == "Parsch"))
capture.output(summary(HSM_M_lmer_Parsch),file = "HeatShock/HSM_M_lmer_sum_Parsch.txt")
capture.output(anova(HSM_M_lmer_Parsch),file = "HeatShock/HSM_M_lmer_Parsch.txt")
capture.output(emmeans(HSM_M_lmer_Parsch, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "HeatShock/HSM_M_lmer_Parsch_tk.txt")

anova(HSM_M_lmer_Parsch)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 1663186  184798     9 46.556  38.121 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(HSM_M_lmer_Parsch)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TimeDeath_min ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: filter(d_HSM_M, Censor == "0" & Supervisor.PI == "Parsch")
## 
## REML criterion at convergence: 57794.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.2106 -0.6839  0.0461  0.6733  2.8035 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 4376     66.15   
##  Batch           (Intercept) 3268     57.17   
##  Residual                    4848     69.63   
## Number of obs: 5065, groups:  Line:Population, 135; Batch, 9
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   382.38      26.00  25.15   14.71 7.34e-14 ***
## PopulationGI   285.06      26.00  25.18   10.96 4.46e-11 ***
## PopulationKA   339.67      25.95  24.97   13.09 1.11e-12 ***
## PopulationMA   387.41      26.00  25.18   14.90 5.37e-14 ***
## PopulationMU   352.29      25.95  24.96   13.58 4.99e-13 ***
## PopulationRE   263.14      25.95  24.95   10.14 2.47e-10 ***
## PopulationUM   369.25      26.25  26.07   14.07 1.09e-13 ***
## PopulationVA   353.08      26.24  26.05   13.45 3.10e-13 ***
## PopulationYE   257.98      25.79  24.40   10.00 4.13e-10 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.543                                                  
## PopulatinKA 0.546  0.552                                           
## PopulatinMA 0.549  0.539  0.545                                    
## PopulatinMU 0.550  0.540  0.546  0.553                             
## PopulatinRE 0.529  0.530  0.531  0.530  0.531                      
## PopulatinUM 0.517  0.521  0.520  0.517  0.518  0.556               
## PopulatinVA 0.517  0.521  0.521  0.517  0.518  0.556  0.560        
## PopulatinYE 0.535  0.538  0.538  0.534  0.536  0.549  0.547  0.547

p_HSM_F_Parsch <- ggplot(data = filter(d_HSM_F, (Censor == "0") & (Supervisor.PI == "Parsch")), aes(x=Population_Lat, y=TimeDeath_min, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_F_Parsch", x="Population", y = "TimeDeath_min") + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_F_Parsch

pdf(file="HeatShock/p_HSM_F_Parsch.pdf",width=8, height=5)
p_HSM_F_Parsch
dev.off()

## quartz_off_screen 
##                 2

p_HSM_F_pop_Parsch <- ggplot(data = filter(d_HSM_F, (Censor == "0") & (Supervisor.PI == "Parsch")), aes(x=Population_Lat, y=TimeDeath_min, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_F_pop_Parsch", x="Population", y = "TimeDeath_min") + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_F_pop_Parsch

pdf(file="HeatShock/p_HSM_F_pop_Parsch.pdf",width=8, height=5)
p_HSM_F_pop_Parsch
dev.off()

## quartz_off_screen 
##                 2

p_HSM_M_Parsch <- ggplot(data = filter(d_HSM_M, (Censor == "0") & (Supervisor.PI == "Parsch")), aes(x=Population_Lat, y=TimeDeath_min, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_M_Parsch", x="Population", y = "TimeDeath_min") + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_M_Parsch

pdf(file="HeatShock/p_HSM_M_Parsch.pdf",width=8, height=5)
p_HSM_M_Parsch
dev.off()

## quartz_off_screen 
##                 2

p_HSM_M_pop_Parsch <- ggplot(data = filter(d_HSM_M, (Censor == "0") & (Supervisor.PI == "Parsch")), aes(x=Population_Lat, y=TimeDeath_min, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_M_pop_Parsch", x="Population", y = "TimeDeath_min") + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_M_pop_Parsch

pdf(file="HeatShock/p_HSM_M_pop_Parsch.pdf",width=8, height=5)
p_HSM_M_pop_Parsch
dev.off()

## quartz_off_screen 
##                 2

Vieira Lab

HSM_F_lmer_Vieira <- lmer(TimeDeath_min ~ -1 + Population  + (1|Line:Population) + (1|Batch), data = filter(d_HSM_F, Censor == "0" & Supervisor.PI == "Vieira"))
capture.output(summary(HSM_F_lmer_Vieira), file = "HeatShock/HSM_F_lmer_sum_Vieira.txt")
capture.output(anova(HSM_F_lmer_Vieira), file = "HeatShock/HSM_F_lmer_Vieira.txt")
capture.output(emmeans(HSM_F_lmer_Vieira, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "HeatShock/HSM_F_lmer_Vieira_tk.txt")

anova(HSM_F_lmer_Vieira)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 97061341 10784593     9 45.874  3538.6 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(HSM_F_lmer_Vieira)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TimeDeath_min ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: filter(d_HSM_F, Censor == "0" & Supervisor.PI == "Vieira")
## 
## REML criterion at convergence: 57881.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -6.2647 -0.5129  0.1397  0.7262  2.3668 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  433.75  20.827  
##  Batch           (Intercept)   17.84   4.223  
##  Residual                    3047.69  55.206  
## Number of obs: 5310, groups:  Line:Population, 168; Batch, 32
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  362.739      5.729  72.734   63.32   <2e-16 ***
## PopulationGI  333.996      6.530  76.988   51.15   <2e-16 ***
## PopulationKA  363.872      6.033  71.727   60.31   <2e-16 ***
## PopulationMA  372.451      5.802  41.351   64.20   <2e-16 ***
## PopulationMU  331.900      5.408  33.797   61.37   <2e-16 ***
## PopulationRE  377.147      6.574  37.291   57.37   <2e-16 ***
## PopulationUM  364.563      5.828  65.949   62.55   <2e-16 ***
## PopulationVA  357.419      5.517  29.304   64.78   <2e-16 ***
## PopulationYE  362.175      5.687  34.993   63.68   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.024                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.005                                    
## PopulatinMU 0.051  0.026  0.001  0.000                             
## PopulatinRE 0.000  0.000  0.007  0.000  0.000                      
## PopulatinUM 0.008  0.010  0.029  0.000  0.022  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.007  0.000  0.000  0.000  0.059

HSM_M_lmer_Vieira <- lmer(TimeDeath_min ~ -1 + Population  + (1|Line:Population) + (1|Batch), data = filter(d_HSM_M, Censor == "0" & Supervisor.PI == "Vieira"))
capture.output(summary(HSM_M_lmer_Vieira), file = "HeatShock/HSM_M_lmer_sum_Vieira.txt")
capture.output(anova(HSM_M_lmer_Vieira), file = "HeatShock/HSM_M_lmer_Vieira.txt")
capture.output(emmeans(HSM_M_lmer_Vieira, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "HeatShock/HSM_M_lmer_Vieira_tk.txt")

anova(HSM_M_lmer_Vieira)

## Type III Analysis of Variance Table with Satterthwaite's method
##              Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 20209997 2245555     9 45.945   753.8 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(HSM_M_lmer_Vieira)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TimeDeath_min ~ -1 + Population + (1 | Line:Population) + (1 |  
##     Batch)
##    Data: filter(d_HSM_M, Censor == "0" & Supervisor.PI == "Vieira")
## 
## REML criterion at convergence: 95195.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.8393 -0.6019  0.0454  0.6858  2.7930 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 1002.0   31.65   
##  Batch           (Intercept)  260.4   16.14   
##  Residual                    2979.0   54.58   
## Number of obs: 8743, groups:  Line:Population, 168; Batch, 32
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  343.597      9.407  61.170   36.53   <2e-16 ***
## PopulationGI  277.186     11.169  75.928   24.82   <2e-16 ***
## PopulationKA  326.787      9.789  52.506   33.38   <2e-16 ***
## PopulationMA  319.605     10.244  37.018   31.20   <2e-16 ***
## PopulationMU  306.202      9.869  58.545   31.03   <2e-16 ***
## PopulationRE  327.221     11.897  30.542   27.50   <2e-16 ***
## PopulationUM  314.506      9.646  64.385   32.61   <2e-16 ***
## PopulationVA  320.016     10.434  36.538   30.67   <2e-16 ***
## PopulationYE  324.995     10.458  38.620   31.08   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.147                                                  
## PopulatinKA 0.008  0.008                                           
## PopulatinMA 0.000  0.000  0.035                                    
## PopulatinMU 0.232  0.133  0.015  0.001                             
## PopulatinRE 0.000  0.000  0.048  0.002  0.001                      
## PopulatinUM 0.071  0.066  0.117  0.004  0.131  0.006               
## PopulatinVA 0.000  0.000  0.000  0.012  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.002  0.045  0.000  0.000  0.000  0.260

p_HSM_F_Vieira <- ggplot(data = filter(d_HSM_F, (Censor == "0") & (Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=TimeDeath_min, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_F_Vieira", x="Population", y = "TimeDeath_min") + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_F_Vieira

pdf(file="HeatShock/p_HSM_F_Vieira.pdf",width=8, height=5)
p_HSM_F_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_HSM_F_pop_Vieira <- ggplot(data = filter(d_HSM_F, (Censor == "0") & (Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=TimeDeath_min, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_F_pop_Vieira", x="Population", y = "TimeDeath_min") + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_F_pop_Vieira

pdf(file="HeatShock/p_HSM_F_pop_Vieira.pdf",width=8, height=5)
p_HSM_F_pop_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_HSM_M_Vieira <- ggplot(data = filter(d_HSM_M, (Censor == "0") & (Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=TimeDeath_min, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_M_Vieira", x="Population", y = "TimeDeath_min") + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_M_Vieira

pdf(file="HeatShock/p_HSM_M_Vieira.pdf",width=8, height=5)
p_HSM_M_Vieira
dev.off()

## quartz_off_screen 
##                 2

p_HSM_M_pop_Vieira <- ggplot(data = filter(d_HSM_M, (Censor == "0") & (Supervisor.PI == "Vieira")), aes(x=Population_Lat, y=TimeDeath_min, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_HSM_M_pop_Vieira", x="Population", y = "TimeDeath_min") + ylim(c(min_HSM, max_HSM))+ droseu_theme

p_HSM_M_pop_Vieira

pdf(file="HeatShock/p_HSM_M_pop_Vieira.pdf",width=8, height=5)
p_HSM_M_pop_Vieira
dev.off()

## quartz_off_screen 
##                 2

Lab correlations

intersect_Line_HSM  <- Reduce(intersect, list(d_HSM$Line[d_HSM$Supervisor.PI == "Parsch"],
                                         d_HSM$Line[d_HSM$Supervisor.PI == "Vieira"]))
write(intersect_Line_HSM, file ="HeatShock/intersect_Line_HSM.txt", sep = "\t")

intersect_Pop_HSM  <- Reduce(intersect, list(d_HSM$Population[d_HSM$Supervisor.PI == "Parsch"],
                                         d_HSM$Population[d_HSM$Supervisor.PI == "Vieira"]))
write(intersect_Pop_HSM, file ="HeatShock/intersect_Pop_HSM.txt", sep = "\t")

There are 9 populations and 132 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

HSM_fitted <- combineFitted(labs = rep(c("Parsch", "Vieira"), 2), sex = rep(c("F", "M"), each = 2), models = list(HSM_F_lmer_Parsch, HSM_F_lmer_Vieira, HSM_M_lmer_Parsch, HSM_M_lmer_Vieira))

p_HSM_F_lab_cor <- function() scatterPlotMatrix(HSM_fitted, sex = "F")
p_HSM_M_lab_cor <- function() scatterPlotMatrix(HSM_fitted, sex = "M")

p_HSM_F_lab_cor()

pdf(file="HeatShock/p_HSM_F_lab_cor.pdf",width=8, height=8)
p_HSM_F_lab_cor()
invisible(dev.off())

p_HSM_M_lab_cor()

pdf(file="HeatShock/p_HSM_M_lab_cor.pdf",width=8, height=8)
p_HSM_M_lab_cor()
invisible(dev.off())

Diapause

dir.create(file.path("Diapause"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Bergland Lab : Liam Miller, Alan Bergland, Priscilla Erickson

Flatt Lab : Esra Durmaz, Envel Kerdaffrec, Thibault Schowing, Virginie Thieu, Marisa Rodrigues, Thomas Flatt

Schlotterer Lab : Manolis Lyrakis, Christian Schlötterer

Data Reformatting

d_Dia <- read.csv("MasterSheets_Oct21_git/DIA_MasterSheet_Oct21.csv")

str(d_Dia)

## 'data.frame':    8206 obs. of  12 variables:
##  $ Supervisor.PI    : chr  "Bergland" "Bergland" "Bergland" "Bergland" ...
##  $ Diet             : chr  "S" "S" "S" "S" ...
##  $ Batch            : int  3 3 3 3 3 3 3 3 3 3 ...
##  $ Population       : chr  "AK" "AK" "AK" "AK" ...
##  $ Line             : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Individual       : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ MostAdvancedStage: int  9 7 10 8 8 8 8 9 9 8 ...
##  $ NumberOfEggs     : int  1 0 3 0 1 3 0 4 3 1 ...
##  $ Country          : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude         : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude        : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude         : int  88 88 88 88 88 88 88 88 88 88 ...

d_Dia <- d_Dia %>%
  mutate(Supervisor.PI = as.factor(Supervisor.PI),
         Diet = as.factor(Diet),
         Batch = as.factor(Batch),
         Population = as.factor(Population),
         Line = as.factor(Line),
         MostAdvancedStage = as.numeric(MostAdvancedStage),
         NumberOfEggs = as.numeric(NumberOfEggs),
         Max_Stage7 = ifelse(MostAdvancedStage <= 7 & NumberOfEggs == 0, 1, 0),
         Max_Stage8 = ifelse(MostAdvancedStage <= 8 & NumberOfEggs == 0, 1, 0),
         Max_Stage9 = ifelse(MostAdvancedStage <= 9 & NumberOfEggs == 0, 1, 0),
         Population_Lat = factor(Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
         Population_Lon = factor(Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA")),
         Population_Alt = factor(Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE")))

str(d_Dia)

## 'data.frame':    8206 obs. of  18 variables:
##  $ Supervisor.PI    : Factor w/ 3 levels "Bergland","Flatt",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet             : Factor w/ 2 levels "NS","S": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Batch            : Factor w/ 3 levels "1","2","3": 3 3 3 3 3 3 3 3 3 3 ...
##  $ Population       : Factor w/ 9 levels "AK","GI","KA",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Line             : Factor w/ 168 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Individual       : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ MostAdvancedStage: num  9 7 10 8 8 8 8 9 9 8 ...
##  $ NumberOfEggs     : num  1 0 3 0 1 3 0 4 3 1 ...
##  $ Country          : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude         : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude        : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude         : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Max_Stage7       : num  0 1 0 0 0 0 0 0 0 0 ...
##  $ Max_Stage8       : num  0 1 0 1 0 0 1 0 0 0 ...
##  $ Max_Stage9       : num  0 1 0 1 0 0 1 0 0 0 ...
##  $ Population_Lat   : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon   : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt   : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

Descriptive Statistics

##### calculate diapause proportions per line, with and without batch
d_Dia_Line_wbatch <- d_Dia %>%
  group_by(Supervisor.PI, Batch, Population, Population_Lat, Population_Lon, Population_Alt, Line) %>%
  dplyr::summarise(n_ind = as.numeric(n()),
            Max_Stage7_Prop = mean(Max_Stage7, na.rm = T),
            Max_Stage8_Prop = mean(Max_Stage8, na.rm = T),
            Max_Stage9_Prop = mean(Max_Stage9, na.rm = T),
            Max_Stage7_Prop_Asin = asin(sqrt(mean(Max_Stage7, na.rm = T))),
            Max_Stage8_Prop_Asin = asin(sqrt(mean(Max_Stage8, na.rm = T))),
            Max_Stage9_Prop_Asin = asin(sqrt(mean(Max_Stage9, na.rm = T))),
            .groups = "drop")
 
d_Dia_Line_wobatch <- d_Dia %>%
  group_by(Supervisor.PI, Population, Population_Lat, Population_Lon, Population_Alt, Line) %>%
  dplyr::summarise(n_ind = as.numeric(n()),
            Max_Stage7_Prop = mean(Max_Stage7, na.rm = T),
            Max_Stage8_Prop = mean(Max_Stage8, na.rm = T),
            Max_Stage9_Prop = mean(Max_Stage9, na.rm = T),
            Max_Stage7_Prop_Asin = asin(sqrt(mean(Max_Stage7, na.rm = T))),
            Max_Stage8_Prop_Asin = asin(sqrt(mean(Max_Stage8, na.rm = T))),
            Max_Stage9_Prop_Asin = asin(sqrt(mean(Max_Stage9, na.rm = T))),
            .groups = "drop")

Descriptive statistics at the line level for Max Stage 9, with batch information :

table_Dia_Line_wbatch <- d_Dia_Line_wbatch %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(Max_Stage9_Prop), list(Mean = mean, SD = sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode))

write.csv(table_Dia_Line_wbatch, "Diapause/table_Dia_Line_wbatch.csv", row.names = T)

Descriptive statistics at the line level for Max Stage 9, without batch information :

table_Dia_Line_wobatch <- d_Dia_Line_wobatch %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(Max_Stage9_Prop), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode))

write.csv(table_Dia_Line_wobatch, "Diapause/table_Dia_Line_wobatch.csv", row.names = T)

Descriptive statistics at the population level, with batch information :

table_Dia_Pop_wbatch <- d_Dia_Line_wbatch %>% group_by(Supervisor.PI, Batch, Population) %>% summarise_at(vars(Max_Stage9_Prop), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode))

write.csv(table_Dia_Pop_wbatch, "Diapause/table_Dia_Pop_wbatch.csv", row.names = T)

Descriptive statistics at the population level, without batch information :

table_Dia_Pop_wobatch <- d_Dia_Line_wobatch %>% group_by(Supervisor.PI, Population) %>% summarise_at(vars(Max_Stage9_Prop), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode))

write.csv(table_Dia_Pop_wobatch, "Diapause/table_Dia_Line_wobatch.csv", row.names = T)

Plots and Linear Models by Lab

Bergland Lab

Dia_lmer_Bergland <- lmer(Max_Stage9_Prop ~ -1 + Population + (1|Population:Line) + (1|Batch) , data = d_Dia_Line_wbatch %>% filter(Supervisor.PI == "Bergland"))
capture.output(summary(Dia_lmer_Bergland),file = "Diapause/Dia_lmer_Bergland_sum.txt")
capture.output(anova(Dia_lmer_Bergland),file = "Diapause/Dia_lmer_Bergland.txt")
capture.output(emmeans(Dia_lmer_Bergland, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Diapause/Dia_lmer_Bergland_tk.txt")

model_means_Bergland <- emmeans(object = Dia_lmer_Bergland, specs = ~ Population)
model_means_Bergland_cld <- cld(object = model_means_Bergland, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

min_Dia <- min(d_Dia_Line_wbatch$Max_Stage9_Prop)
max_Dia <- max(d_Dia_Line_wbatch$Max_Stage9_Prop)

p_Dia_Pop_Bergland_wbatch <- d_Dia_Line_wbatch %>% 
  filter(Supervisor.PI == 'Bergland') %>%
  ggplot(aes(x = Population_Lat, y = Max_Stage9_Prop, fill = Batch)) + 
  geom_boxplot(outlier.shape = NA, notch = FALSE) +  
  labs(title = "p_Dia_Bergland", x = "Population", y = "Max_Stage9_Prop") +
  droseu_theme +
  ylim(c(min_Dia, max_Dia))
p_Dia_Pop_Bergland_wbatch

pdf(file="Diapause/p_Dia_Pop_Bergland_wbatch.pdf",width=8, height=5)
p_Dia_Pop_Bergland_wbatch
dev.off()

## quartz_off_screen 
##                 2

p_Dia_Pop_Bergland_wobatch <- d_Dia_Line_wobatch %>% 
  filter(Supervisor.PI == 'Bergland') %>%
  inner_join(model_means_Bergland_cld) %>%
  ggplot(aes(x = Population_Lat, y = Max_Stage9_Prop, fill = Population)) + 
  geom_boxplot(outlier.shape = NA, notch = FALSE) +  
  geom_text(data = model_means_Bergland_cld, aes(x = Population, label = cld), y = 1, vjust = -1) +
  labs(title = "p_Dia_Bergland", x = "Population", y = "Max_Stage9_Prop") +
  droseu_theme +
  ylim(c(min_Dia, max_Dia))

## Joining, by = "Population"

p_Dia_Pop_Bergland_wobatch

pdf(file="Diapause/p_Dia_Pop_Bergland_wobatch.pdf",width=8, height=5)
p_Dia_Pop_Bergland_wobatch
dev.off()

## quartz_off_screen 
##                 2

Flatt Lab

Dia_lm_Flatt <- lm(Max_Stage9_Prop ~ -1 + Population, data = d_Dia_Line_wbatch %>% filter(Supervisor.PI == "Flatt"))
capture.output(summary(Dia_lm_Flatt),file = "Diapause/Dia_lm_Flatt_sum.txt")
capture.output(anova(Dia_lm_Flatt),file = "Diapause/Dia_lm_Flatt")
capture.output(emmeans(Dia_lm_Flatt, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Diapause/Dia_lm_Flatt_tk.txt")

model_means_Flatt <- emmeans(object = Dia_lm_Flatt, specs = ~ Population)
model_means_Flatt_cld <- cld(object = model_means_Flatt, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

p_Dia_Pop_Flatt_wbatch <- d_Dia_Line_wbatch %>% 
  filter(Supervisor.PI == 'Flatt') %>%
  ggplot(aes(x = Population_Lat, y = Max_Stage9_Prop, fill = Batch)) + 
  geom_boxplot(outlier.shape = NA, notch = FALSE) +  
  labs(title = "p_Dia_Flatt", x = "Population", y = "Max_Stage9_Prop") +
  droseu_theme +
  ylim(c(min_Dia, max_Dia))
p_Dia_Pop_Flatt_wbatch

pdf(file="Diapause/p_Dia_Pop_Flatt_wbatch.pdf", width=8, height=5)
p_Dia_Pop_Flatt_wbatch
dev.off()

## quartz_off_screen 
##                 2

p_Dia_Pop_Flatt_wobatch <- d_Dia_Line_wobatch %>% 
  filter(Supervisor.PI == 'Flatt') %>%
  ggplot(aes(x = Population_Lat, y = Max_Stage9_Prop, fill = Population)) + 
  geom_boxplot(outlier.shape = NA, notch = FALSE) +  
  geom_text(data = model_means_Flatt_cld, aes(x = Population, label = cld), y = 1, vjust = -1) +
  labs(title = "p_Dia_Flatt", x = "Population", y = "Max_Stage9_Prop") +
  droseu_theme +
  ylim(c(min_Dia, max_Dia))
p_Dia_Pop_Flatt_wobatch

pdf(file="Diapause/p_Dia_Pop_Flatt_wobatch.pdf", width=8, height=5)
p_Dia_Pop_Flatt_wobatch
dev.off()

## quartz_off_screen 
##                 2

Schlotterer Lab

Dia_lmer_Schlotterer <- lmer(Max_Stage9_Prop ~ -1 + Population + (1|Population:Line) + (1|Batch) , data = d_Dia_Line_wbatch %>% filter(Supervisor.PI == "Schlotterer"))
capture.output(summary(Dia_lmer_Schlotterer),file = "Diapause/Dia_lmer_Schlotterer_sum.txt")
capture.output(anova(Dia_lmer_Schlotterer),file = "Diapause/Dia_lmer_Schlotterer.txt")
capture.output(emmeans(Dia_lmer_Schlotterer, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Diapause/Dia_lmer_Schlotterer_tk.txt")

model_means_Schlotterer <- emmeans(object = Dia_lmer_Schlotterer, specs = ~ Population)
model_means_Schlotterer_cld <- cld(object = model_means_Schlotterer, adjust = "Tukey", Letters = letters, alpha = 0.05) %>% as.data.frame %>% mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))) %>% arrange(Population) %>% dplyr::rename(cld = .group)

## Note: adjust = "tukey" was changed to "sidak"
## because "tukey" is only appropriate for one set of pairwise comparisons

p_Dia_Pop_Schlotterer_wbatch <- d_Dia_Line_wbatch %>% 
  filter(Supervisor.PI == 'Schlotterer') %>%
  ggplot(aes(x = Population_Lat, y = Max_Stage9_Prop, fill = Batch)) + 
  geom_boxplot(outlier.shape = NA, notch = FALSE) +  
  labs(title = "p_Dia_Schlotterer", x = "Population", y = "Max_Stage9_Prop") +
  droseu_theme +
  ylim(c(min_Dia, max_Dia))
p_Dia_Pop_Schlotterer_wbatch

pdf(file="Diapause/p_Dia_Pop_Schlotterer_wbatch.pdf", width=8, height=5)
p_Dia_Pop_Schlotterer_wbatch
dev.off()

## quartz_off_screen 
##                 2

p_Dia_Pop_Schlotterer_wobatch <- d_Dia_Line_wobatch %>% 
  filter(Supervisor.PI == 'Schlotterer') %>%
  ggplot(aes(x = Population_Lat, y = Max_Stage9_Prop, fill = Population)) + 
  geom_boxplot(outlier.shape = NA, notch = FALSE) +  
  geom_text(data = model_means_Schlotterer_cld, aes(x = Population, label = cld), y = 1, vjust = -0.7) +
  labs(title = "p_Dia_Schlotterer", x = "Population", y = "Max_Stage9_Prop") +
  droseu_theme +
  ylim(c(min_Dia, max_Dia))
p_Dia_Pop_Schlotterer_wobatch

pdf(file="Diapause/p_Dia_Pop_Schlotterer_wobatch.pdf", width=8, height=5)
p_Dia_Pop_Schlotterer_wobatch
dev.off()

## quartz_off_screen 
##                 2

Lab correlations

intersect_Line_Dia  <- Reduce(intersect, list(d_Dia$Line[d_Dia$Supervisor.PI == "Flatt"],
                                         d_Dia$Line[d_Dia$Supervisor.PI == "Bergland"],
                                         d_Dia$Line[d_Dia$Supervisor.PI == "Schlotterer"]))

write(intersect_Line_Dia, file ="Diapause/intersect_Line_Dia.txt", sep = "\t")

intersect_Pop_Dia  <- Reduce(intersect, list(d_Dia$Population[d_Dia$Supervisor.PI == "Flatt"],
                                         d_Dia$Population[d_Dia$Supervisor.PI == "Bergland"],
                                         d_Dia$Population[d_Dia$Supervisor.PI == "Schlotterer"]))

write(intersect_Pop_Dia, file ="Diapause/intersect_Pop_Dia.txt", sep = "\t")

There are 9 populations and 146 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

Dia_fitted <- combineFitted(labs = c("Bergland", "Flatt", "Schlotterer"), sex = rep("F", 3), models = list(Dia_lmer_Bergland, Dia_lm_Flatt, Dia_lmer_Schlotterer))

p_Dia_F_lab_cor <- function() scatterPlotMatrix(Dia_fitted, sex = "F")

p_Dia_F_lab_cor()

pdf(file="Diapause/p_Dia_F_lab_cor.pdf",width=8, height=8)
p_Dia_F_lab_cor()
invisible(dev.off())

Circadian eclosion timing

dir.create(file.path("CircadianEclosion"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

#Note that the trait has been phenotyped as mixed sex. 

Contributors

Wegener Lab : Susanne Klühspies, Christian Wegener

Data Reformatting

d_CET <- read.csv("MasterSheets_Oct21_git/CET_MasterSheet_Dec21.csv")

str(d_CET)

## 'data.frame':    36 obs. of  16 variables:
##  $ Supervisor.PI                 : chr  "Wegener" "Wegener" "Wegener" "Wegener" ...
##  $ Diet                          : chr  "NS" "NS" "NS" "NS" ...
##  $ total...of.flies              : int  1897 793 1938 1355 2349 1418 1178 1544 2150 1236 ...
##  $ Batch                         : int  2 2 2 2 3 2 1 2 3 2 ...
##  $ Population                    : chr  "AK" "GI" "KA" "MA" ...
##  $ Line                          : chr  "mixed population" "mixed population" "mixed population" "mixed population" ...
##  $ Sex                           : chr  "mixed sex" "mixed sex" "mixed sex" "mixed sex" ...
##  $ ReplicateVial                 : logi  NA NA NA NA NA NA ...
##  $ Individual                    : logi  NA NA NA NA NA NA ...
##  $ Condition                     : chr  "18_LD_DD" "18_LD_DD" "18_LD_DD" "18_LD_DD" ...
##  $ CET_hours_MESA                : num  13.2 10.5 14 16.8 10.5 ...
##  $ CET_hours_LSPR                : num  12.5 10.5 11.7 17.5 11 ...
##  $ Period_MESA                   : num  24.2 24.8 24.1 23.3 24.9 ...
##  $ Period_LSPR                   : num  24.4 24.8 24.8 23.1 24.7 ...
##  $ Rhythmicity_LSPR_amplitude    : num  11.66 6.55 9.75 7.47 16.88 ...
##  $ Rhythmicity_JTK_p_BH_corrected: num  5.26e-24 6.27e-22 6.76e-14 2.92e-09 4.21e-24 ...

d_CET$Supervisor.PI <- as.factor(d_CET$Supervisor.PI)
d_CET$Diet <- as.factor(d_CET$Diet)
d_CET$Batch <- as.factor(d_CET$Batch)
d_CET$Population_Lat <- factor(d_CET$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_CET$Population_Lon <- factor(d_CET$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_CET$Population_Alt <- factor(d_CET$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_CET$Condition <- as.factor(d_CET$Condition)
d_CET$CET_hours_MESA <- as.numeric(d_CET$CET_hours_MESA)
d_CET$CET_hours_LSPR <- as.numeric(d_CET$CET_hours_LSPR)
d_CET$Period_MESA <- as.numeric(d_CET$Period_MESA)
d_CET$Period_LSPR <- as.numeric(d_CET$Period_LSPR)
d_CET$Rhythmicity_LSPR_amplitude <- as.numeric(d_CET$Rhythmicity_LSPR_amplitude)
d_CET$Rhythmicity_JTK_p_BH_corrected <- as.numeric(d_CET$Rhythmicity_JTK_p_BH_corrected)

str(d_CET)

## 'data.frame':    36 obs. of  19 variables:
##  $ Supervisor.PI                 : Factor w/ 1 level "Wegener": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet                          : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ total...of.flies              : int  1897 793 1938 1355 2349 1418 1178 1544 2150 1236 ...
##  $ Batch                         : Factor w/ 4 levels "1","2","3","4": 2 2 2 2 3 2 1 2 3 2 ...
##  $ Population                    : chr  "AK" "GI" "KA" "MA" ...
##  $ Line                          : chr  "mixed population" "mixed population" "mixed population" "mixed population" ...
##  $ Sex                           : chr  "mixed sex" "mixed sex" "mixed sex" "mixed sex" ...
##  $ ReplicateVial                 : logi  NA NA NA NA NA NA ...
##  $ Individual                    : logi  NA NA NA NA NA NA ...
##  $ Condition                     : Factor w/ 4 levels "18_LD_DD","18_LD_LD",..: 1 1 1 1 1 1 1 1 1 2 ...
##  $ CET_hours_MESA                : num  13.2 10.5 14 16.8 10.5 ...
##  $ CET_hours_LSPR                : num  12.5 10.5 11.7 17.5 11 ...
##  $ Period_MESA                   : num  24.2 24.8 24.1 23.3 24.9 ...
##  $ Period_LSPR                   : num  24.4 24.8 24.8 23.1 24.7 ...
##  $ Rhythmicity_LSPR_amplitude    : num  11.66 6.55 9.75 7.47 16.88 ...
##  $ Rhythmicity_JTK_p_BH_corrected: num  5.26e-24 6.27e-22 6.76e-14 2.92e-09 4.21e-24 ...
##  $ Population_Lat                : Factor w/ 9 levels "YE","RE","GI",..: 9 3 7 5 4 2 6 8 1 9 ...
##  $ Population_Lon                : Factor w/ 9 levels "RE","GI","KA",..: 6 2 3 5 4 1 7 9 8 6 ...
##  $ Population_Alt                : Factor w/ 9 levels "KA","AK","GI",..: 2 3 1 8 7 4 5 6 9 2 ...

d_CET_18_LD_DD <-subset(d_CET,Condition=='18_LD_DD')

d_CET_18_LD_LD <-subset(d_CET,Condition=='18_LD_LD')

d_CET_29_LD_DD <-subset(d_CET,Condition=='29_LD_DD')

d_CET_29_LD_LD <-subset(d_CET,Condition=='29_LD_LD')

Descriptive Statistics

18_LD_DD

Descriptive statistics for CET_hours_MESA :

table_CET_18LDDD_CET_MESA <- write.csv(d_CET_18_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                    summarise_at(vars(CET_hours_MESA), 
                                                 list(Mean = mean)), 
                            file = "CircadianEclosion/table_CET_18LDDD_CET_MESA.csv", row.names = T)
table_CET_18LDDD_CET_MESA <- read.csv("CircadianEclosion/table_CET_18LDDD_CET_MESA.csv")

DT::datatable(table_CET_18LDDD_CET_MESA, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for CET_hours_LSPR :

table_CET_18LDDD_CET_LSPR <- write.csv(d_CET_18_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                    summarise_at(vars(CET_hours_LSPR), 
                                                 list(Mean = mean)), 
                            file = "CircadianEclosion/table_CET_18LDDD_CET_LSPR.csv", row.names = T)
table_CET_18LDDD_CET_LSPR <- read.csv("CircadianEclosion/table_CET_18LDDD_CET_LSPR.csv")

DT::datatable(table_CET_18LDDD_CET_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Period_MESA :

table_CET_18LDDD_Period_MESA <- write.csv(d_CET_18_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                    summarise_at(vars(Period_MESA), 
                                                 list(Mean = mean)), 
                            file = "CircadianEclosion/table_CET_18LDDD_Period_MESA.csv", row.names = T)
table_CET_18LDDD_Period_MESA <- read.csv("CircadianEclosion/table_CET_18LDDD_Period_MESA.csv")

DT::datatable(table_CET_18LDDD_Period_MESA, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Period_LSPR :

table_CET_18LDDD_Period_LSPR <- write.csv(d_CET_18_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                    summarise_at(vars(Period_LSPR), 
                                                 list(Mean = mean)), 
                            file = "CircadianEclosion/table_CET_18LDDD_Period_LSPR.csv", row.names = T)
table_CET_18LDDD_Period_LSPR <- read.csv("CircadianEclosion/table_CET_18LDDD_Period_LSPR.csv")

DT::datatable(table_CET_18LDDD_Period_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Rhythmicity_LSPR_amplitude :

table_CET_18LDDD_Rhy_LSPR <- write.csv(d_CET_18_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                    summarise_at(vars(Rhythmicity_LSPR_amplitude), 
                                                 list(Mean = mean)), 
                            file = "CircadianEclosion/table_CET_18LDDD_Rhy_LSPR.csv", row.names = T)
table_CET_18LDDD_Rhy_LSPR <- read.csv("CircadianEclosion/table_CET_18LDDD_Rhy_LSPR.csv")

DT::datatable(table_CET_18LDDD_Rhy_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Rhythmicity_JTK_p_BH_corrected :

table_CET_18LDDD_Rhy_JTK <- write.csv(d_CET_18_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                    summarise_at(vars(Rhythmicity_JTK_p_BH_corrected), 
                                                 list(Mean = mean)), 
                            file = "CircadianEclosion/table_CET_18LDDD_Rhy_JTK.csv", row.names = T)
table_CET_18LDDD_Rhy_JTK <- read.csv("CircadianEclosion/table_CET_18LDDD_Rhy_JTK.csv")

DT::datatable(table_CET_18LDDD_Rhy_JTK, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

18_LD_LD

Descriptive statistics for CET_hours_MESA :

table_CET_18LDLD_CET_MESA <- write.csv(d_CET_18_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(CET_hours_MESA), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_18LDLD_CET_MESA.csv", row.names = T)
table_CET_18LDLD_CET_MESA <- read.csv("CircadianEclosion/table_CET_18LDLD_CET_MESA.csv")

DT::datatable(table_CET_18LDLD_CET_MESA, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for CET_hours_LSPR :

table_CET_18LDLD_CET_LSPR <- write.csv(d_CET_18_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(CET_hours_LSPR), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_18LDLD_CET_LSPR.csv", row.names = T)
table_CET_18LDLD_CET_LSPR <- read.csv("CircadianEclosion/table_CET_18LDLD_CET_LSPR.csv")

DT::datatable(table_CET_18LDLD_CET_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Period_MESA :

table_CET_18LDLD_Period_MESA <- write.csv(d_CET_18_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                            summarise_at(vars(Period_MESA), 
                                                         list(Mean = mean)), 
                                          file = "CircadianEclosion/table_CET_18LDLD_Period_MESA.csv", row.names = T)
table_CET_18LDLD_Period_MESA <- read.csv("CircadianEclosion/table_CET_18LDLD_Period_MESA.csv")

DT::datatable(table_CET_18LDLD_Period_MESA, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Period_LSPR :

table_CET_18LDLD_Period_LSPR <- write.csv(d_CET_18_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                            summarise_at(vars(Period_LSPR), 
                                                         list(Mean = mean)), 
                                          file = "CircadianEclosion/table_CET_18LDLD_Period_LSPR.csv", row.names = T)
table_CET_18LDLD_Period_LSPR <- read.csv("CircadianEclosion/table_CET_18LDLD_Period_LSPR.csv")

DT::datatable(table_CET_18LDLD_Period_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Rhythmicity_LSPR_amplitude :

table_CET_18LDLD_Rhy_LSPR <- write.csv(d_CET_18_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(Rhythmicity_LSPR_amplitude), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_18LDLD_Rhy_LSPR.csv", row.names = T)
table_CET_18LDLD_Rhy_LSPR <- read.csv("CircadianEclosion/table_CET_18LDLD_Rhy_LSPR.csv")

DT::datatable(table_CET_18LDLD_Rhy_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Rhythmicity_JTK_p_BH_corrected :

table_CET_18LDLD_Rhy_JTK <- write.csv(d_CET_18_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                        summarise_at(vars(Rhythmicity_JTK_p_BH_corrected), 
                                                     list(Mean = mean)), 
                                      file = "CircadianEclosion/table_CET_18LDLD_Rhy_JTK.csv", row.names = T)
table_CET_18LDLD_Rhy_JTK <- read.csv("CircadianEclosion/table_CET_18LDLD_Rhy_JTK.csv")

DT::datatable(table_CET_18LDLD_Rhy_JTK, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

29_LD_DD

Descriptive statistics for CET_hours_MESA :

table_CET_29LDDD_CET_MESA <- write.csv(d_CET_29_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(CET_hours_MESA), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_29LDDD_CET_MESA.csv", row.names = T)
table_CET_29LDDD_CET_MESA <- read.csv("CircadianEclosion/table_CET_29LDDD_CET_MESA.csv")

DT::datatable(table_CET_29LDDD_CET_MESA, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for CET_hours_LSPR :

table_CET_29LDDD_CET_LSPR <- write.csv(d_CET_29_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(CET_hours_LSPR), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_29LDDD_CET_LSPR.csv", row.names = T)
table_CET_29LDDD_CET_LSPR <- read.csv("CircadianEclosion/table_CET_29LDDD_CET_LSPR.csv")

DT::datatable(table_CET_29LDDD_CET_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Period_MESA :

table_CET_29LDDD_Period_MESA <- write.csv(d_CET_29_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                            summarise_at(vars(Period_MESA), 
                                                         list(Mean = mean)), 
                                          file = "CircadianEclosion/table_CET_29LDDD_Period_MESA.csv", row.names = T)
table_CET_29LDDD_Period_MESA <- read.csv("CircadianEclosion/table_CET_29LDDD_Period_MESA.csv")

DT::datatable(table_CET_29LDDD_Period_MESA, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Period_LSPR :

table_CET_29LDDD_Period_LSPR <- write.csv(d_CET_29_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                            summarise_at(vars(Period_LSPR), 
                                                         list(Mean = mean)), 
                                          file = "CircadianEclosion/table_CET_29LDDD_Period_LSPR.csv", row.names = T)
table_CET_29LDDD_Period_LSPR <- read.csv("CircadianEclosion/table_CET_29LDDD_Period_LSPR.csv")

DT::datatable(table_CET_29LDDD_Period_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Rhythmicity_LSPR_amplitude :

table_CET_29LDDD_Rhy_LSPR <- write.csv(d_CET_29_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(Rhythmicity_LSPR_amplitude), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_29LDDD_Rhy_LSPR.csv", row.names = T)
table_CET_29LDDD_Rhy_LSPR <- read.csv("CircadianEclosion/table_CET_29LDDD_Rhy_LSPR.csv")

DT::datatable(table_CET_29LDDD_Rhy_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Rhythmicity_JTK_p_BH_corrected :

table_CET_29LDDD_Rhy_JTK <- write.csv(d_CET_29_LD_DD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                        summarise_at(vars(Rhythmicity_JTK_p_BH_corrected), 
                                                     list(Mean = mean)), 
                                      file = "CircadianEclosion/table_CET_29LDDD_Rhy_JTK.csv", row.names = T)
table_CET_29LDDD_Rhy_JTK <- read.csv("CircadianEclosion/table_CET_29LDDD_Rhy_JTK.csv")

DT::datatable(table_CET_29LDDD_Rhy_JTK, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

29_LD_LD

Descriptive statistics for CET_hours_MESA :

table_CET_29LDLD_CET_MESA <- write.csv(d_CET_29_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(CET_hours_MESA), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_29LDLD_CET_MESA.csv", row.names = T)
table_CET_29LDLD_CET_MESA <- read.csv("CircadianEclosion/table_CET_29LDLD_CET_MESA.csv")

DT::datatable(table_CET_29LDLD_CET_MESA, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for CET_hours_LSPR :

table_CET_29LDLD_CET_LSPR <- write.csv(d_CET_29_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(CET_hours_LSPR), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_29LDLD_CET_LSPR.csv", row.names = T)
table_CET_29LDLD_CET_LSPR <- read.csv("CircadianEclosion/table_CET_29LDLD_CET_LSPR.csv")

DT::datatable(table_CET_29LDLD_CET_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Period_MESA :

table_CET_29LDLD_Period_MESA <- write.csv(d_CET_29_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                            summarise_at(vars(Period_MESA), 
                                                         list(Mean = mean)), 
                                          file = "CircadianEclosion/table_CET_29LDLD_Period_MESA.csv", row.names = T)
table_CET_29LDLD_Period_MESA <- read.csv("CircadianEclosion/table_CET_29LDLD_Period_MESA.csv")

DT::datatable(table_CET_29LDLD_Period_MESA, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Period_LSPR :

table_CET_29LDLD_Period_LSPR <- write.csv(d_CET_29_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                            summarise_at(vars(Period_LSPR), 
                                                         list(Mean = mean)), 
                                          file = "CircadianEclosion/table_CET_29LDLD_Period_LSPR.csv", row.names = T)
table_CET_29LDLD_Period_LSPR <- read.csv("CircadianEclosion/table_CET_29LDLD_Period_LSPR.csv")

DT::datatable(table_CET_29LDLD_Period_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Rhythmicity_LSPR_amplitude :

table_CET_29LDLD_Rhy_LSPR <- write.csv(d_CET_29_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                         summarise_at(vars(Rhythmicity_LSPR_amplitude), 
                                                      list(Mean = mean)), 
                                       file = "CircadianEclosion/table_CET_29LDLD_Rhy_LSPR.csv", row.names = T)
table_CET_29LDLD_Rhy_LSPR <- read.csv("CircadianEclosion/table_CET_29LDLD_Rhy_LSPR.csv")

DT::datatable(table_CET_29LDLD_Rhy_LSPR, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Descriptive statistics for Rhythmicity_JTK_p_BH_corrected :

table_CET_29LDLD_Rhy_JTK <- write.csv(d_CET_29_LD_LD %>% group_by(Supervisor.PI, Condition, Population) %>% 
                                        summarise_at(vars(Rhythmicity_JTK_p_BH_corrected), 
                                                     list(Mean = mean)), 
                                      file = "CircadianEclosion/table_CET_29LDLD_Rhy_JTK.csv", row.names = T)
table_CET_29LDLD_Rhy_JTK <- read.csv("CircadianEclosion/table_CET_29LDLD_Rhy_JTK.csv")

DT::datatable(table_CET_29LDLD_Rhy_JTK, options = list(pageLength = 10)) %>% formatRound(5:5, 3)

Plots and Linear Models by Lab

Overview

Wegener Lab

From Chris :

knitr::include_graphics("CircadianEclosion/FromChris/Graphs_plate_MESA.png")

From Eran :

knitr::include_graphics("CircadianEclosion/FromEran/eclosion_data_Christian_Page_1.png")

knitr::include_graphics("CircadianEclosion/FromEran/eclosion_data_Christian_Page_2.png")

Locomotor activity

dir.create(file.path("Locomotor"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

#Note that the trait has been phenotyped only in males.

Contributors

Tauber Lab : Bettina Fishman, Eran Tauber

Data Reformatting

d_LA <- read.csv("MasterSheets_Oct21_git/LA_MasterSheet_Oct21.csv")

str(d_LA)

## 'data.frame':    639 obs. of  19 variables:
##  $ Supervisor.PI      : chr  "Tauber" "Tauber" "Tauber" "Tauber" ...
##  $ Diet               : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch              : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population         : chr  "YE" "YE" "YE" "YE" ...
##  $ Line               : chr  "YE27" "YE33" "YE40" "YE40" ...
##  $ Data.Label         : chr  "YE 29-9-18" "YE 33-9-18" "YE 40-9-18" "YE 40-9-18" ...
##  $ Sex                : chr  "M" "M" "M" "M" ...
##  $ ReplicateChamberOld: int  3 3 3 3 3 3 3 3 3 3 ...
##  $ ReplicateChamber   : chr  "Tauber_1_YE27_M_3" "Tauber_1_YE33_M_3" "Tauber_1_YE40_M_3" "Tauber_1_YE40_M_3" ...
##  $ Period             : num  23.6 20.2 24 24.2 23.1 ...
##  $ CircPhase          : num  23.21 13.07 4.59 9.32 12.98 ...
##  $ AbsPhase           : num  12.5 0.7 11.7 12.5 12.5 12 12.7 0.5 12.4 12.6 ...
##  $ Experiment         : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ND                 : num  0.678 0.95 1.083 0.948 1.455 ...
##  $ Activity           : num  29.2 46.8 45.4 37.8 37.8 31.8 43.6 42.8 29.6 21.6 ...
##  $ Country            : chr  "Turkey" "Turkey" "Turkey" "Turkey" ...
##  $ Latitude           : num  40.2 40.2 40.2 40.2 40.2 ...
##  $ Longitude          : num  32.3 32.3 32.3 32.3 32.3 ...
##  $ Altitude           : int  680 680 680 680 680 680 680 680 680 680 ...

d_LA$Supervisor.PI <- as.factor(d_LA$Supervisor.PI)
d_LA$Diet <- as.factor(d_LA$Diet)
d_LA$Batch <- as.factor(d_LA$Batch)
d_LA$Population_Lat <- factor(d_LA$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_LA$Population_Lon <- factor(d_LA$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_LA$Population_Alt <- factor(d_LA$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_LA$Line <- as.factor(d_LA$Line)
d_LA$ReplicateChamber <- as.factor(d_LA$ReplicateChamber)
d_LA$Period <- as.numeric(d_LA$Period)
d_LA$CircPhase <- as.numeric(d_LA$CircPhase)
d_LA$AbsPhase <- as.numeric(d_LA$AbsPhase)
d_LA$Experiment <- as.numeric(d_LA$Experiment)
d_LA$ND <- as.numeric(d_LA$ND)
d_LA$Activity <- as.numeric(d_LA$Activity)

str(d_LA)

## 'data.frame':    639 obs. of  22 variables:
##  $ Supervisor.PI      : Factor w/ 1 level "Tauber": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet               : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch              : Factor w/ 2 levels "1","2": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population         : chr  "YE" "YE" "YE" "YE" ...
##  $ Line               : Factor w/ 92 levels "AK1","AK2","AK3",..: 84 85 86 86 86 89 90 90 91 91 ...
##  $ Data.Label         : chr  "YE 29-9-18" "YE 33-9-18" "YE 40-9-18" "YE 40-9-18" ...
##  $ Sex                : chr  "M" "M" "M" "M" ...
##  $ ReplicateChamberOld: int  3 3 3 3 3 3 3 3 3 3 ...
##  $ ReplicateChamber   : Factor w/ 152 levels "Tauber_1_AK1_M_9",..: 76 77 78 78 78 79 80 80 81 81 ...
##  $ Period             : num  23.6 20.2 24 24.2 23.1 ...
##  $ CircPhase          : num  23.21 13.07 4.59 9.32 12.98 ...
##  $ AbsPhase           : num  12.5 0.7 11.7 12.5 12.5 12 12.7 0.5 12.4 12.6 ...
##  $ Experiment         : num  1 1 1 1 1 1 1 1 1 1 ...
##  $ ND                 : num  0.678 0.95 1.083 0.948 1.455 ...
##  $ Activity           : num  29.2 46.8 45.4 37.8 37.8 31.8 43.6 42.8 29.6 21.6 ...
##  $ Country            : chr  "Turkey" "Turkey" "Turkey" "Turkey" ...
##  $ Latitude           : num  40.2 40.2 40.2 40.2 40.2 ...
##  $ Longitude          : num  32.3 32.3 32.3 32.3 32.3 ...
##  $ Altitude           : int  680 680 680 680 680 680 680 680 680 680 ...
##  $ Population_Lat     : Factor w/ 9 levels "YE","RE","GI",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population_Lon     : Factor w/ 9 levels "RE","GI","KA",..: 8 8 8 8 8 8 8 8 8 8 ...
##  $ Population_Alt     : Factor w/ 9 levels "KA","AK","GI",..: 9 9 9 9 9 9 9 9 9 9 ...

Descriptive Statistics

LA - Activity

Descriptive statistics at the line level, with batch information :

table_LA_Act_Line_wbatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Batch, Sex, Population, Line) %>% 
                                    summarise_at(vars(Activity), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "Locomotor/table_LA_Act_Line_wbatch.csv", row.names = T)
table_LA_Act_Line_wbatch <- read.csv("Locomotor/table_LA_Act_Line_wbatch.csv")

Descriptive statistics at the line level, without batch information :

table_LA_Act_Line_wobatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Sex, Population, Line) %>% 
                                    summarise_at(vars(Activity), 
                                                 list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                      SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                            file = "Locomotor/table_LA_Act_Line_wobatch.csv", row.names = T)
table_LA_Act_Line_wobatch <- read.csv("Locomotor/table_LA_Act_Line_wobatch.csv")

Descriptive statistics at the population level, with batch information :

table_LA_Act_Pop_wbatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Batch, Sex, Population) %>% 
                                   summarise_at(vars(Activity), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Locomotor/table_LA_Act_Pop_wbatch.csv", row.names = T)
table_LA_Act_Pop_wbatch <- read.csv("Locomotor/table_LA_Act_Pop_wbatch.csv")

Descriptive statistics at the population level, without batch information :

table_LA_Act_Pop_wobatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Sex, Population) %>% 
                                   summarise_at(vars(Activity), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Locomotor/table_LA_Act_Pop_wobatch.csv", row.names = T)
table_LA_Act_Pop_wobatch <- read.csv("Locomotor/table_LA_Act_Pop_wobatch.csv")

LA - ND ratio

Descriptive statistics at the line level, with batch information :

table_LA_ND_Line_wbatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Batch, Sex, Population, Line) %>% 
                                        summarise_at(vars(ND), 
                                                     list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                          SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                                      file = "Locomotor/table_LA_ND_Line_wbatch.csv", row.names = T)
table_LA_ND_Line_wbatch <- read.csv("Locomotor/table_LA_ND_Line_wbatch.csv")

Descriptive statistics at the line level, without batch information :

table_LA_ND_Line_wobatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Sex, Population, Line) %>% 
                                         summarise_at(vars(ND), 
                                                      list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                           SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                                       file = "Locomotor/table_LA_ND_Line_wobatch.csv", row.names = T)
table_LA_ND_Line_wobatch <- read.csv("Locomotor/table_LA_ND_Line_wobatch.csv")

Descriptive statistics at the population level, with batch information :

table_LA_ND_Pop_wbatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Batch, Sex, Population) %>% 
                                       summarise_at(vars(ND), 
                                                    list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                         SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                                     file = "Locomotor/table_LA_ND_Pop_wbatch.csv", row.names = T)
table_LA_ND_Pop_wbatch <- read.csv("Locomotor/table_LA_ND_Pop_wbatch.csv")

Descriptive statistics at the population level, without batch information :

table_LA_ND_Pop_wobatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Sex, Population) %>% 
                                        summarise_at(vars(ND), 
                                                     list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                          SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                                      file = "Locomotor/table_LA_ND_Pop_wobatch.csv", row.names = T)
table_LA_ND_Pop_wobatch <- read.csv("Locomotor/table_LA_ND_Pop_wobatch.csv")

LA - Circadian Phase

Descriptive statistics at the line level, with batch information :

table_LA_CircPhase_Line_wbatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Batch, Sex, Population, Line) %>% 
                                        summarise_at(vars(CircPhase), 
                                                     list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                          SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                                      file = "Locomotor/table_LA_CircPhase_Line_wbatch.csv", row.names = T)
table_LA_CircPhase_Line_wbatch <- read.csv("Locomotor/table_LA_CircPhase_Line_wbatch.csv")

Descriptive statistics at the line level, without batch information :

table_LA_CircPhase_Line_wobatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Sex, Population, Line) %>% 
                                         summarise_at(vars(CircPhase), 
                                                      list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                           SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                                       file = "Locomotor/table_LA_CircPhase_Line_wobatch.csv", row.names = T)
table_LA_CircPhase_Line_wobatch <- read.csv("Locomotor/table_LA_CircPhase_Line_wobatch.csv")

Descriptive statistics at the population level, with batch information :

table_LA_CircPhase_Pop_wbatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Batch, Sex, Population) %>% 
                                       summarise_at(vars(CircPhase), 
                                                    list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                         SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                                     file = "Locomotor/table_LA_CircPhase_Pop_wbatch.csv", row.names = T)
table_LA_CircPhase_Pop_wbatch <- read.csv("Locomotor/table_LA_CircPhase_Pop_wbatch.csv")

Descriptive statistics at the population level, without batch information :

table_LA_CircPhase_Pop_wobatch <- write.csv(d_LA %>% group_by(Supervisor.PI, Sex, Population) %>% 
                                        summarise_at(vars(CircPhase), 
                                                     list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                          SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                                      file = "Locomotor/table_LA_CircPhase_Pop_wobatch.csv", row.names = T)
table_LA_CircPhase_Pop_wobatch <- read.csv("Locomotor/table_LA_CircPhase_Pop_wobatch.csv")

Plots and Linear Models by Lab

Overview

Tauber Lab

ND

LA_lmer_ND_Tauber <- lmer(ND ~ -1 + Population  + (1|Line:Population) +
                          (1|Batch), data = (d_LA))

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00214269 (tol = 0.002, component 1)

capture.output(summary(LA_lmer_ND_Tauber),file = "Locomotor/LA_lmer_ND_Tauber_sum.txt")
capture.output(anova(LA_lmer_ND_Tauber),file = "Locomotor/LA_lmer_ND_Tauber.txt")
capture.output(emmeans(LA_lmer_ND_Tauber, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Locomotor/LA_lmer_ND_Tauber_tk.txt")

anova(LA_lmer_ND_Tauber)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population  182.2  20.245     9 74.675  136.84 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LA_lmer_ND_Tauber)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: ND ~ -1 + Population + (1 | Line:Population) + (1 | Batch)
##    Data: (d_LA)
## 
## REML criterion at convergence: 703.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.1922 -0.4257 -0.0668  0.2481 11.4146 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev. 
##  Line:Population (Intercept) 3.972e-02 0.1993100
##  Batch           (Intercept) 1.608e-09 0.0000401
##  Residual                    1.479e-01 0.3846341
## Number of obs: 639, groups:  Line:Population, 92; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  0.86109    0.08409 63.09959  10.240 4.83e-15 ***
## PopulationGI  0.85162    0.08587 75.56622   9.918 2.52e-15 ***
## PopulationKA  1.02359    0.07759 67.28287  13.193  < 2e-16 ***
## PopulationMA  0.95795    0.07734 67.48095  12.386  < 2e-16 ***
## PopulationMU  0.86577    0.08728 95.77068   9.919 2.27e-16 ***
## PopulationRE  0.88107    0.09339 84.94375   9.435 7.17e-15 ***
## PopulationUM  0.80411    0.08216 69.94382   9.787 9.80e-15 ***
## PopulationVA  0.91036    0.08277 66.44809  10.998  < 2e-16 ***
## PopulationYE  1.08890    0.06321 98.77181  17.226  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## Model failed to converge with max|grad| = 0.00214269 (tol = 0.002, component 1)

CircPhase

LA_lmer_CircPhase_Tauber <- lmer(CircPhase ~ -1 + Population  + (1|Line:Population) +
                          (1|Batch) , data = (d_LA))
capture.output(summary(LA_lmer_CircPhase_Tauber),file = "Locomotor/LA_lmer_CircPhase_Tauber_sum.txt")
capture.output(anova(LA_lmer_CircPhase_Tauber),file = "Locomotor/LA_lmer_CircPhase_Tauber.txt")
capture.output(emmeans(LA_lmer_CircPhase_Tauber, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Locomotor/LA_lmer_CircPhase_Tauber_tk.txt")

anova(LA_lmer_CircPhase_Tauber)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value  Pr(>F)  
## Population  26773  2974.8     9     2  63.875 0.01551 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LA_lmer_CircPhase_Tauber)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: CircPhase ~ -1 + Population + (1 | Line:Population) + (1 | Batch)
##    Data: (d_LA)
## 
## REML criterion at convergence: 4247.6
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.10384 -0.57625 -0.06697  0.85374  1.83691 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  0.04096 0.2024  
##  Batch           (Intercept)  0.41358 0.6431  
##  Residual                    46.57281 6.8244  
## Number of obs: 639, groups:  Line:Population, 92; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   9.0667     0.9362  8.2604   9.684 8.62e-06 ***
## PopulationGI  13.2114     0.9879 10.3499  13.374 7.30e-08 ***
## PopulationKA  14.0128     0.8646  6.1489  16.206 2.82e-06 ***
## PopulationMA  12.9574     0.8807  6.6656  14.712 2.50e-06 ***
## PopulationMU  12.1721     1.0796 14.7993  11.275 1.16e-08 ***
## PopulationRE  13.2907     1.1150 13.9621  11.920 1.05e-08 ***
## PopulationUM  10.9990     0.9320  8.2560  11.801 1.87e-06 ***
## PopulationVA  12.9768     0.9065  7.2307  14.316 1.44e-06 ***
## PopulationYE  12.9227     0.8484  5.9741  15.231 5.25e-06 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.224                                                  
## PopulatinKA 0.253  0.242                                           
## PopulatinMA 0.245  0.237  0.273                                    
## PopulatinMU 0.210  0.194  0.220  0.214                             
## PopulatinRE 0.176  0.187  0.219  0.224  0.161                      
## PopulatinUM 0.234  0.225  0.257  0.254  0.204  0.205               
## PopulatinVA 0.236  0.231  0.265  0.264  0.208  0.221  0.247        
## PopulatinYE 0.253  0.246  0.284  0.281  0.222  0.235  0.263  0.274

Activity

LA_lmer_Activity_Tauber <- lmer(Activity ~ -1 + Population  + (1|Line:Population) +
                          (1|Batch), data = (d_LA))
capture.output(summary(LA_lmer_Activity_Tauber),file = "Locomotor/LA_lmer_Activity_Tauber_sum.txt")
capture.output(anova(LA_lmer_Activity_Tauber),file = "Locomotor/LA_lmer_Activity_Tauber.txt")
capture.output(emmeans(LA_lmer_Activity_Tauber, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Locomotor/LA_lmer_Activity_Tauber_tk.txt")

anova(LA_lmer_Activity_Tauber)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value Pr(>F)  
## Population  19295  2143.9     9     2   31.03 0.0316 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(LA_lmer_Activity_Tauber)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: Activity ~ -1 + Population + (1 | Line:Population) + (1 | Batch)
##    Data: (d_LA)
## 
## REML criterion at convergence: 4624.6
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -3.16030 -0.55380  0.07697  0.65610  2.31656 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 41.72    6.459   
##  Batch           (Intercept)  7.27    2.696   
##  Residual                    69.09    8.312   
## Number of obs: 639, groups:  Line:Population, 92; Batch, 2
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   31.197      3.140  6.396   9.935 4.02e-05 ***
## PopulationGI   34.315      3.129  6.353  10.967 2.31e-05 ***
## PopulationKA   35.781      2.968  5.186  12.057 5.45e-05 ***
## PopulationMA   36.842      2.964  5.160  12.430 4.83e-05 ***
## PopulationMU   25.126      3.109  6.236   8.081 0.000158 ***
## PopulationRE   38.221      3.299  7.751  11.586 3.62e-06 ***
## PopulationUM   25.960      3.066  5.865   8.468 0.000167 ***
## PopulationVA   28.507      3.087  6.020   9.235 8.93e-05 ***
## PopulationYE   34.247      2.606  3.152  13.142 0.000738 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.370                                                  
## PopulatinKA 0.390  0.391                                           
## PopulatinMA 0.389  0.392  0.413                                    
## PopulatinMU 0.373  0.374  0.394  0.394                             
## PopulatinRE 0.345  0.351  0.371  0.375  0.352                      
## PopulatinUM 0.377  0.379  0.400  0.400  0.381  0.361               
## PopulatinVA 0.373  0.376  0.397  0.398  0.378  0.361  0.385        
## PopulatinYE 0.442  0.446  0.470  0.472  0.448  0.428  0.456  0.453

From Eran :

Red letters denote significant difference by Multiple Comparisons of Means: Tukey Contrasts

knitr::include_graphics("Locomotor/FromEran/pheno2022_Page_1.png")

knitr::include_graphics("Locomotor/FromEran/pheno2022_Page_2.png")

knitr::include_graphics("Locomotor/FromEran/pheno2022_Page_3.png")

Phase in DD represented in angle. The mean and 95 confidence limits are shown. CI limits in red indicate uniform distribution (low concentration. i.e. Rayleigh test not significant).

knitr::include_graphics("Locomotor/FromEran/pheno2022_Page_4.png")

Oriana 3 (Stacked circular).

knitr::include_graphics("Locomotor/FromEran/pheno2022_Page_5.png")

Starvation resistance

dir.create(file.path("Starvation"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Gonzalez Lab : Llewellyn Green, Josefa Gonzalez, Miriam Merenciano

Onder Lab : Seda Coskun, Ekin Demir, Senel Selin Senkal, Cansu Aksoy

Pasyukova Lab : Alexander Symonenko, Natalia Roshina, Mikhail Trostnokov, Ekaterina Veselkina, Evgenia Tsybul’ko, Olga Rybina, Elena Pasyukova

Data Reformatting

d_SR <- read.csv("MasterSheets_Oct21_git/SR_MasterSheet_Nov21.csv")

str(d_SR)

## 'data.frame':    78315 obs. of  14 variables:
##  $ Supervisor.PI   : chr  "Gonzalez" "Gonzalez" "Gonzalez" "Gonzalez" ...
##  $ Diet            : chr  "S" "S" "S" "S" ...
##  $ Batch           : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Sex             : chr  "F" "F" "F" "F" ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : chr  "Gonzalez_1_AK1_F_1" "Gonzalez_1_AK1_F_1" "Gonzalez_1_AK1_F_1" "Gonzalez_1_AK1_F_1" ...
##  $ Individual      : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ AgeAtDeath_hours: int  56 96 96 96 104 104 104 112 112 120 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...

d_SR$Supervisor.PI <- as.factor(d_SR$Supervisor.PI)
d_SR$Diet <- as.factor(d_SR$Diet)
d_SR$Batch <- as.factor(d_SR$Batch)
d_SR$Population_Lat <- factor(d_SR$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_SR$Population_Lon <- factor(d_SR$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_SR$Population_Alt <- factor(d_SR$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_SR$Line <- as.factor(d_SR$Line)
d_SR$Sex <- as.factor(d_SR$Sex)
d_SR$ReplicateVial <- as.factor(d_SR$ReplicateVial)
d_SR$AgeAtDeath_hours <- as.numeric(d_SR$AgeAtDeath_hours)
str(d_SR)

## 'data.frame':    78315 obs. of  17 variables:
##  $ Supervisor.PI   : Factor w/ 3 levels "Gonzalez","Onder",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet            : Factor w/ 2 levels "NS","S": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Batch           : Factor w/ 7 levels "1","2","3","4",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population      : chr  "AK" "AK" "AK" "AK" ...
##  $ Line            : Factor w/ 171 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex             : Factor w/ 2 levels "F","M": 1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVialOld: int  1 1 1 1 1 1 1 1 1 1 ...
##  $ ReplicateVial   : Factor w/ 7753 levels "Gonzalez_1_AK1_F_1",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Individual      : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ AgeAtDeath_hours: num  56 96 96 96 104 104 104 112 112 120 ...
##  $ Country         : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude        : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude       : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude        : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat  : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon  : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt  : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_SR_F <-subset(d_SR,Sex=='F')
d_SR_M <-subset(d_SR,Sex=='M')

Descriptive Statistics

Descriptive statistics at the line level, with batch information :

table_SR_Line_wbatch <- write.csv(d_SR %>% group_by(Supervisor.PI, Batch, Sex,Population, Line) %>% 
                                   summarise_at(vars(AgeAtDeath_hours), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Starvation/table_SR_Line_wbatch.csv", row.names = T)
table_SR_Line_wbatch <- read.csv("Starvation/table_SR_Line_wbatch.csv")

Descriptive statistics at the line level, without batch information :

table_SR_Line_wobatch <- write.csv(d_SR %>% group_by(Supervisor.PI, Sex,Population, Line) %>% 
                                   summarise_at(vars(AgeAtDeath_hours), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Starvation/table_SR_Line_wobatch.csv", row.names = T)
table_SR_Line_wobatch <- read.csv("Starvation/table_SR_Line_wobatch.csv")

Descriptive statistics at the population level, with batch information :

table_SR_Pop_wbatch <- write.csv(d_SR %>% group_by(Supervisor.PI, Batch, Sex, Population) %>% 
                                  summarise_at(vars(AgeAtDeath_hours), 
                                               list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                    SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                          file = "Starvation/table_SR_Pop_wbatch.csv", row.names = T)

table_SR_Pop_wbatch <- read.csv("Starvation/table_SR_Pop_wbatch.csv")

Descriptive statistics at the population level, without batch information :

table_SR_Pop_wobatch <- write.csv(d_SR %>% group_by(Supervisor.PI, Sex, Population) %>% 
                                  summarise_at(vars(AgeAtDeath_hours), 
                                               list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                    SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                          file = "Starvation/table_SR_Pop_wobatch.csv", row.names = T)

table_SR_Pop_wobatch <- read.csv("Starvation/table_SR_Pop_wobatch.csv")

Plots and Linear Models by Lab

Overview

min_SR <- min(d_SR$AgeAtDeath_hours)
max_SR <- max(d_SR$AgeAtDeath_hours)

p_SR_F <- ggplot(data = d_SR_F, aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + labs(title="p_SR_F", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_M <- ggplot(data = d_SR_M, aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Supervisor.PI)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + labs(title="p_SR_M", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

Gonzalez Lab

SR_F_lmer_Gonzalez <- lmer(AgeAtDeath_hours ~ -1 + Population  + (1|Line:Population) +
                          (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_SR_F,Supervisor.PI=='Gonzalez')))
capture.output(summary(SR_F_lmer_Gonzalez),file = "Starvation/SR_F_lmer_Gonzalez_sum.txt")
capture.output(anova(SR_F_lmer_Gonzalez),file = "Starvation/SR_F_lmer_Gonzalez.txt")
capture.output(emmeans(SR_F_lmer_Gonzalez, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Starvation/SR_F_lmer_Gonzalez_tk.txt")

anova(SR_F_lmer_Gonzalez)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 704125   78236     9 56.752  231.46 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(SR_F_lmer_Gonzalez)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: AgeAtDeath_hours ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_SR_F, Supervisor.PI == "Gonzalez"))
## 
## REML criterion at convergence: 43270.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.6657 -0.6316 -0.0456  0.5852  4.9362 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept) 117.17   10.824  
##  Line:Population               (Intercept) 126.21   11.234  
##  Batch                         (Intercept)  11.45    3.385  
##  Residual                                  338.01   18.385  
## Number of obs: 4900, groups:  
## ReplicateVial:Line:Population, 436; Line:Population, 141; Batch, 7
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   85.162      3.288  71.182   25.90   <2e-16 ***
## PopulationGI   74.698      3.914 105.480   19.08   <2e-16 ***
## PopulationKA   80.832      3.773 101.847   21.42   <2e-16 ***
## PopulationMA   79.731      3.262  76.797   24.44   <2e-16 ***
## PopulationMU   88.602      3.388  75.107   26.15   <2e-16 ***
## PopulationRE   79.754      3.906 106.983   20.42   <2e-16 ***
## PopulationUM   81.717      4.203 107.764   19.44   <2e-16 ***
## PopulationVA   88.398      3.616  83.263   24.44   <2e-16 ***
## PopulationYE   69.534      3.750  88.007   18.54   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.149                                                  
## PopulatinKA 0.148  0.119                                           
## PopulatinMA 0.183  0.145  0.144                                    
## PopulatinMU 0.194  0.148  0.145  0.183                             
## PopulatinRE 0.142  0.116  0.116  0.141  0.142                      
## PopulatinUM 0.149  0.116  0.114  0.142  0.149  0.112               
## PopulatinVA 0.183  0.138  0.136  0.170  0.182  0.131  0.138        
## PopulatinYE 0.181  0.136  0.133  0.168  0.181  0.129  0.137  0.170

SR_M_lmer_Gonzalez <- lmer(AgeAtDeath_hours ~ -1 + Population  + (1|Line:Population) +
                                   (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_SR_M,Supervisor.PI=='Gonzalez')))
capture.output(summary(SR_M_lmer_Gonzalez),file = "Starvation/SR_M_lmer_Gonzalez_sum.txt")
capture.output(anova(SR_M_lmer_Gonzalez),file = "Starvation/SR_M_lmer_Gonzalez.txt")
capture.output(emmeans(SR_M_lmer_Gonzalez, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Starvation/SR_M_lmer_Gonzalez_tk.txt")

anova(SR_M_lmer_Gonzalez)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population  45603    5067     9 34.009  33.912 1.958e-14 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(SR_M_lmer_Gonzalez)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: AgeAtDeath_hours ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_SR_M, Supervisor.PI == "Gonzalez"))
## 
## REML criterion at convergence: 42786.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.8769 -0.6117 -0.0757  0.5087  8.1754 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  49.61    7.043  
##  Line:Population               (Intercept)  60.30    7.766  
##  Batch                         (Intercept)  79.31    8.905  
##  Residual                                  149.42   12.224  
## Number of obs: 5335, groups:  
## ReplicateVial:Line:Population, 481; Line:Population, 156; Batch, 7
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   59.290      3.992 11.231   14.85 9.84e-09 ***
## PopulationGI   51.077      4.115 12.673   12.41 1.84e-08 ***
## PopulationKA   59.083      4.103 12.533   14.40 3.62e-09 ***
## PopulationMA   53.956      3.946 10.758   13.67 3.85e-08 ***
## PopulationMU   64.822      4.006 11.386   16.18 3.27e-09 ***
## PopulationRE   54.826      4.097 12.472   13.38 9.11e-09 ***
## PopulationUM   57.365      4.222 13.979   13.59 1.91e-09 ***
## PopulationVA   57.791      3.944 10.734   14.65 1.95e-08 ***
## PopulationYE   47.775      3.939 10.686   12.13 1.39e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.704                                                  
## PopulatinKA 0.702  0.678                                           
## PopulatinMA 0.734  0.708  0.707                                    
## PopulatinMU 0.732  0.704  0.700  0.734                             
## PopulatinRE 0.691  0.672  0.674  0.701  0.690                      
## PopulatinUM 0.691  0.666  0.663  0.694  0.692  0.655               
## PopulatinVA 0.736  0.709  0.708  0.740  0.735  0.701  0.695        
## PopulatinYE 0.733  0.708  0.707  0.738  0.732  0.702  0.693  0.739

p_SR_F_Gonzalez <- ggplot(data = (subset(d_SR_F,Supervisor.PI=='Gonzalez')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_F_Gonzalez", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_F_pop_Gonzalez <- ggplot(data = (subset(d_SR_F,Supervisor.PI=='Gonzalez')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_F_pop_Gonzalez", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_M_Gonzalez <- ggplot(data = (subset(d_SR_M,Supervisor.PI=='Gonzalez')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_M_Gonzalez", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_M_pop_Gonzalez <- ggplot(data = (subset(d_SR_M,Supervisor.PI=='Gonzalez')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_M_pop_Gonzalez", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

Onder Lab

SR_F_lmer_Onder <- lmer(AgeAtDeath_hours ~ -1 + Population  + (1|Line:Population) +
                                (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_SR_F,Supervisor.PI=='Onder')))

## boundary (singular) fit: see ?isSingular

## Warning: Model failed to converge with 1 negative eigenvalue: -6.7e+00

capture.output(summary(SR_F_lmer_Onder),file = "Starvation/SR_F_lmer_Onder_sum.txt")
capture.output(anova(SR_F_lmer_Onder),file = "Starvation/SR_F_lmer_Onder.txt")
capture.output(emmeans(SR_F_lmer_Onder, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Starvation/SR_F_lmer_Onder_tk.txt")

anova(SR_F_lmer_Onder)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2677788  297532     9 158.95  948.75 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(SR_F_lmer_Onder)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: AgeAtDeath_hours ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_SR_F, Supervisor.PI == "Onder"))
## 
## REML criterion at convergence: 149434.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.0413 -0.6188 -0.0307  0.6086  3.9849 
## 
## Random effects:
##  Groups                        Name        Variance  Std.Dev. 
##  ReplicateVial:Line:Population (Intercept) 9.720e+00 3.118e+00
##  Line:Population               (Intercept) 1.683e+02 1.297e+01
##  Batch                         (Intercept) 3.004e-07 5.481e-04
##  Residual                                  3.136e+02 1.771e+01
## Number of obs: 17282, groups:  
## ReplicateVial:Line:Population, 1737; Line:Population, 168; Batch, 5
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   92.651      2.936 159.004   31.56   <2e-16 ***
## PopulationGI   88.175      3.390 159.025   26.01   <2e-16 ***
## PopulationKA   97.242      2.936 158.998   33.12   <2e-16 ***
## PopulationMA   92.373      2.936 159.020   31.46   <2e-16 ***
## PopulationMU  102.780      2.935 158.705   35.02   <2e-16 ***
## PopulationRE   88.775      3.282 158.821   27.05   <2e-16 ***
## PopulationUM   93.081      3.185 159.123   29.22   <2e-16 ***
## PopulationVA  102.602      2.935 158.816   34.96   <2e-16 ***
## PopulationYE   80.139      2.936 159.003   27.30   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

SR_M_lmer_Onder <- lmer(AgeAtDeath_hours ~ -1 + Population  + (1|Line:Population) +
                                (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_SR_M,Supervisor.PI=='Onder')))
capture.output(summary(SR_M_lmer_Onder),file = "Starvation/SR_M_lmer_Onder_sum.txt")
capture.output(anova(SR_M_lmer_Onder),file = "Starvation/SR_M_lmer_Onder.txt")
capture.output(emmeans(SR_M_lmer_Onder, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Starvation/SR_M_lmer_Onder_tk.txt")

anova(SR_M_lmer_Onder)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 1299875  144431     9 88.625  774.35 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(SR_M_lmer_Onder)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: AgeAtDeath_hours ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_SR_M, Supervisor.PI == "Onder"))
## 
## REML criterion at convergence: 139728.1
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.1862 -0.6258 -0.0367  0.5970  5.2151 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept)  10.5535  3.2486 
##  Line:Population               (Intercept)  95.9375  9.7948 
##  Batch                         (Intercept)   0.8965  0.9468 
##  Residual                                  186.5178 13.6572 
## Number of obs: 17158, groups:  
## ReplicateVial:Line:Population, 1732; Line:Population, 168; Batch, 5
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   73.073      2.271 131.418   32.18   <2e-16 ***
## PopulationGI   69.646      2.616 141.305   26.63   <2e-16 ***
## PopulationKA   75.445      2.267 134.598   33.28   <2e-16 ***
## PopulationMA   73.215      2.268 133.692   32.29   <2e-16 ***
## PopulationMU   79.813      2.269 131.295   35.17   <2e-16 ***
## PopulationRE   70.052      2.530 140.185   27.69   <2e-16 ***
## PopulationUM   76.321      2.453 141.600   31.11   <2e-16 ***
## PopulationVA   77.707      2.265 135.176   34.31   <2e-16 ***
## PopulationYE   64.875      2.265 135.195   28.64   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.035                                                  
## PopulatinKA 0.033  0.033                                           
## PopulatinMA 0.039  0.034  0.034                                    
## PopulatinMU 0.041  0.034  0.032  0.039                             
## PopulatinRE 0.037  0.032  0.030  0.036  0.037                      
## PopulatinUM 0.030  0.029  0.035  0.030  0.029  0.027               
## PopulatinVA 0.038  0.033  0.034  0.037  0.038  0.034  0.031        
## PopulatinYE 0.038  0.034  0.035  0.037  0.038  0.035  0.032  0.036

p_SR_F_Onder <- ggplot(data = (subset(d_SR_F,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_F_Onder", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_F_pop_Onder <- ggplot(data = (subset(d_SR_F,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_F_pop_Onder", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_M_Onder <- ggplot(data = (subset(d_SR_M,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_M_Onder", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_M_pop_Onder <- ggplot(data = (subset(d_SR_M,Supervisor.PI=='Onder')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_M_pop_Onder", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

Pasyukova Lab

SR_F_lmer_Pasyukova <- lmer(AgeAtDeath_hours ~ -1 + Population  + (1|Line:Population) +
                                    (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_SR_F,Supervisor.PI=='Pasyukova')))

## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00451614 (tol = 0.002, component 1)

capture.output(summary(SR_F_lmer_Pasyukova),file = "Starvation/SR_F_lmer_Pasyukova_sum.txt")
capture.output(anova(SR_F_lmer_Pasyukova),file = "Starvation/SR_F_lmer_Pasyukova.txt")
capture.output(emmeans(SR_F_lmer_Pasyukova, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Starvation/SR_F_lmer_Pasyukova_tk.txt")

anova(SR_F_lmer_Pasyukova)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 260966   28996     9 11.49  77.817 5.843e-09 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(SR_F_lmer_Pasyukova)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: AgeAtDeath_hours ~ -1 + Population + (1 | Line:Population) +  
##     (1 | Batch) + (1 | ReplicateVial:Line:Population)
##    Data: (subset(d_SR_F, Supervisor.PI == "Pasyukova"))
## 
## REML criterion at convergence: 149510.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.1750 -0.5760  0.0268  0.6125  3.5353 
## 
## Random effects:
##  Groups                        Name        Variance Std.Dev.
##  ReplicateVial:Line:Population (Intercept) 101.25   10.062  
##  Line:Population               (Intercept) 113.33   10.646  
##  Batch                         (Intercept)  73.51    8.574  
##  Residual                                  372.62   19.303  
## Number of obs: 16779, groups:  
## ReplicateVial:Line:Population, 1681; Line:Population, 169; Batch, 5
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK  103.165      4.621   5.335   22.32 1.78e-06 ***
## PopulationGI  108.267      5.022   7.451   21.56 5.51e-08 ***
## PopulationKA  105.116      4.674   5.597   22.49 1.05e-06 ***
## PopulationMA  101.321      4.683   5.623   21.64 1.25e-06 ***
## PopulationMU  105.822      4.683   5.626   22.59 9.73e-07 ***
## PopulationRE  103.531      4.849   6.462   21.35 3.10e-07 ***
## PopulationUM  103.944      4.759   5.994   21.84 6.08e-07 ***
## PopulationVA  110.685      4.683   5.624   23.64 7.61e-07 ***
## PopulationYE   88.479      4.676   5.606   18.92 2.70e-06 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.668                                                  
## PopulatinKA 0.719  0.659                                           
## PopulatinMA 0.719  0.659  0.709                                    
## PopulatinMU 0.719  0.659  0.709  0.710                             
## PopulatinRE 0.695  0.636  0.685  0.685  0.685                      
## PopulatinUM 0.708  0.649  0.698  0.698  0.699  0.675               
## PopulatinVA 0.719  0.659  0.709  0.710  0.710  0.685  0.698        
## PopulatinYE 0.719  0.658  0.708  0.709  0.709  0.685  0.698  0.709 
## optimizer (nloptwrap) convergence code: 0 (OK)
## Model failed to converge with max|grad| = 0.00451614 (tol = 0.002, component 1)

SR_M_lmer_Pasyukova <- lmer(AgeAtDeath_hours ~ -1 + Population  + (1|Line:Population) +
                                    (1|Batch) + (1|ReplicateVial : Line : Population), data = (subset(d_SR_M,Supervisor.PI=='Pasyukova')))
capture.output(summary(SR_M_lmer_Pasyukova),file = "Starvation/SR_M_lmer_Pasyukova_sum.txt")
capture.output(anova(SR_M_lmer_Pasyukova),file = "Starvation/SR_M_lmer_Pasyukova.txt")
capture.output(emmeans(SR_M_lmer_Pasyukova, list(pairwise ~ Population), adjust = "tukey", mode = "asymp"),file = "Starvation/SR_M_lmer_Pasyukova_tk.txt")

p_SR_F_Pasyukova <- ggplot(data = (subset(d_SR_F,Supervisor.PI=='Pasyukova')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_F_Pasyukova", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_F_pop_Pasyukova <- ggplot(data = (subset(d_SR_F,Supervisor.PI=='Pasyukova')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_F_pop_Pasyukova", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_M_Pasyukova <- ggplot(data = (subset(d_SR_M,Supervisor.PI=='Pasyukova')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_M_Pasyukova", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

p_SR_M_pop_Pasyukova <- ggplot(data = (subset(d_SR_M,Supervisor.PI=='Pasyukova')), aes(x=Population_Lat, y=AgeAtDeath_hours, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_SR_M_pop_Pasyukova", x="Population", y = "AgeAtDeath_hours") + ylim(c(min_SR, max_SR))+ droseu_theme

## quartz_off_screen 
##                 2

Survival Analyses

d_SR_surv <- d_SR %>% mutate(Censor = 1)

Gonzalez Lab

Females

SR_F_Gonzalez_survfit <- survfit(Surv(AgeAtDeath_hours) ~ Population, data = filter(d_SR_surv, Supervisor.PI == "Gonzalez", Sex == "F"))

p_SR_F_Gonzalez_survfit <- ggsurvplot(SR_F_Gonzalez_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "p_SR_F_Gonzalez")

p_SR_F_Gonzalez_survfit$plot

pdf(file="Starvation/p_SR_F_Gonzalez_survfit.pdf", width=8, height=5)
p_SR_F_Gonzalez_survfit$plot
invisible(dev.off())

Males

SR_M_Gonzalez_survfit <- survfit(Surv(AgeAtDeath_hours) ~ Population, data = filter(d_SR_surv, Supervisor.PI == "Gonzalez", Sex == "M"))

p_SR_M_Gonzalez_survfit <- ggsurvplot(SR_M_Gonzalez_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "p_SR_M_Gonzalez")

p_SR_M_Gonzalez_survfit$plot

pdf(file="Starvation/p_SR_M_Gonzalez_survfit.pdf", width=8, height=5)
p_SR_M_Gonzalez_survfit$plot
invisible(dev.off())

Onder Lab

Females

SR_F_Onder_survfit <- survfit(Surv(AgeAtDeath_hours) ~ Population, data = filter(d_SR_surv, Supervisor.PI == "Onder", Sex == "F"))

p_SR_F_Onder_survfit <- ggsurvplot(SR_F_Onder_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "p_SR_F_Onder")

p_SR_F_Onder_survfit$plot

pdf(file="Starvation/p_SR_F_Onder_survfit.pdf", width=8, height=5)
p_SR_F_Onder_survfit$plot
invisible(dev.off())

Males

SR_M_Onder_survfit <- survfit(Surv(AgeAtDeath_hours) ~ Population, data = filter(d_SR_surv, Supervisor.PI == "Onder", Sex == "M"))

p_SR_M_Onder_survfit <- ggsurvplot(SR_M_Onder_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "p_SR_M_Onder")

p_SR_M_Onder_survfit$plot

pdf(file="Starvation/p_SR_M_Onder_survfit.pdf", width=8, height=5)
p_SR_M_Onder_survfit$plot
invisible(dev.off())

Pasyukova Lab

Females

SR_F_Pasyukova_survfit <- survfit(Surv(AgeAtDeath_hours) ~ Population, data = filter(d_SR_surv, Supervisor.PI == "Pasyukova", Sex == "F"))

p_SR_F_Pasyukova_survfit <- ggsurvplot(SR_F_Pasyukova_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "p_SR_F_Pasyukova")

p_SR_F_Pasyukova_survfit$plot

pdf(file="Starvation/p_SR_F_Pasyukova_survfit.pdf", width=8, height=5)
p_SR_F_Pasyukova_survfit$plot
invisible(dev.off())

Males

SR_M_Pasyukova_survfit <- survfit(Surv(AgeAtDeath_hours) ~ Population, data = filter(d_SR_surv, Supervisor.PI == "Pasyukova", Sex == "M"))

p_SR_M_Pasyukova_survfit <- ggsurvplot(SR_M_Pasyukova_survfit, censor.shape="|", censor.size = 0, pval = FALSE, risk.table = FALSE, conf.int = FALSE, legend.title = "Population", legend.labs = sort(c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), legend = c(0.9, 0.7), title = "p_SR_M_Pasyukova")

p_SR_M_Pasyukova_survfit$plot

pdf(file="Starvation/p_SR_M_Pasyukova_survfit.pdf", width=8, height=5)
p_SR_M_Pasyukova_survfit$plot
invisible(dev.off())

coxme

SR_F_coxme_Gonzalez <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_SR_surv, Supervisor.PI == "Gonzalez", Sex == "F"))
capture.output(summary(SR_F_coxme_Gonzalez), file = "Starvation/SR_F_coxme_Gonzalez_sum.txt")
capture.output(anova(SR_F_coxme_Gonzalez), file = "Starvation/SR_F_coxme_Gonzalez.txt")

SR_M_coxme_Gonzalez <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_SR_surv, Supervisor.PI == "Gonzalez", Sex == "M"))
capture.output(summary(SR_M_coxme_Gonzalez), file = "Starvation/SR_M_coxme_Gonzalez_sum.txt")
capture.output(anova(SR_M_coxme_Gonzalez), file = "Starvation/SR_M_coxme_Gonzalez.txt")

SR_F_coxme_Onder <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_SR_surv, Supervisor.PI == "Onder", Sex == "F"))
capture.output(summary(SR_F_coxme_Onder), file = "Starvation/SR_F_coxme_Onder_sum.txt")
capture.output(anova(SR_F_coxme_Onder), file = "Starvation/SR_F_coxme_Onder.txt")

SR_M_coxme_Onder <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_SR_surv, Supervisor.PI == "Onder", Sex == "M"))
capture.output(summary(SR_M_coxme_Onder), file = "Starvation/SR_M_coxme_Onder_sum.txt")
capture.output(anova(SR_M_coxme_Onder), file = "Starvation/SR_M_coxme_Onder.txt")

SR_F_coxme_Pasyukova <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_SR_surv, Supervisor.PI == "Pasyukova", Sex == "F"))
capture.output(summary(SR_F_coxme_Pasyukova), file = "Starvation/SR_F_coxme_Pasyukova_sum.txt")
capture.output(anova(SR_F_coxme_Pasyukova), file = "Starvation/SR_F_coxme_Pasyukova.txt")

SR_M_coxme_Pasyukova <- coxme(Surv(AgeAtDeath_hours, Censor) ~ Population + (1|Batch) + (1|Population/Line) + (1|Population/Line/ReplicateVial), data = filter(d_SR_surv, Supervisor.PI == "Pasyukova", Sex == "M"))
capture.output(summary(SR_M_coxme_Pasyukova), file = "Starvation/SR_M_coxme_Pasyukova_sum.txt")
capture.output(anova(SR_M_coxme_Pasyukova), file = "Starvation/SR_M_coxme_Pasyukova.txt")

Lab correlations

intersect_Line_SR  <- Reduce(intersect, list(d_SR$Line[d_SR$Supervisor.PI == "Gonzalez"],
                                        d_SR$Line[d_SR$Supervisor.PI == "Pasyukova"],
                                        d_SR$Line[d_SR$Supervisor.PI == "Onder"]))

write(intersect_Line_SR, file ="Starvation/intersect_Line_SR.txt", sep = "\t")

intersect_Pop_SR  <- Reduce(intersect, list(d_SR$Population[d_SR$Supervisor.PI == "Gonzalez"],
                                        d_SR$Population[d_SR$Supervisor.PI == "Pasyukova"],
                                        d_SR$Line[d_SR$Supervisor.PI == "Onder"]))
write(intersect_Pop_SR, file ="Starvation/intersect_Pop_SR.txt", sep = "\t")

There are 0 populations and 156 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

SR_fitted <- combineFitted(labs = rep(c("Gonzalez", "Onder", "Pasyukova"), 2), sex = rep(c("F", "M"), each = 3), models = list(SR_F_lmer_Gonzalez, SR_F_lmer_Onder, SR_F_lmer_Pasyukova, SR_M_lmer_Gonzalez, SR_M_lmer_Onder, SR_M_lmer_Pasyukova))

p_SR_F_lab_cor <- function() scatterPlotMatrix(SR_fitted, sex = "F")
p_SR_M_lab_cor <- function() scatterPlotMatrix(SR_fitted, sex = "M")

p_SR_F_lab_cor()

p_SR_M_lab_cor()

pdf(file="Starvation/p_SR_F_lab_cor.pdf", width=8, height=8)
p_SR_F_lab_cor()
invisible(dev.off())

pdf(file="Starvation/p_SR_M_lab_cor.pdf", width=8, height=8)
p_SR_M_lab_cor()
invisible(dev.off())

Pigmentation

dir.create(file.path("Pigmentation"), showWarnings = FALSE)

For a detailed description of tables, plots, linear models and outputs, please refer here

Contributors

Abbott Lab : Jessica Abbott, Qinyang Li, Shahzad Khan

Gibert Lab :Cristina Vieira, Laurence Mouton, Natacha Kremer, Sonia Martinez, Camille Mermet, Patricia Gibert

Schmidt Lab : Amy Goldfischer, Paul Schmidt

Data Reformatting

d_Pgm <- read.csv("MasterSheets_Oct21_git/PGM_MasterSheet_Jan22.csv")

str(d_Pgm)

## 'data.frame':    3694 obs. of  17 variables:
##  $ Supervisor.PI: chr  "Gibert" "Gibert" "Gibert" "Gibert" ...
##  $ Diet         : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch        : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population   : chr  "AK" "AK" "AK" "AK" ...
##  $ Line         : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Sex          : logi  FALSE FALSE FALSE FALSE FALSE FALSE ...
##  $ Individual   : int  1 2 3 4 5 6 7 8 9 1 ...
##  $ AreaT4       : int  12439 10230 7525 14898 9160 8279 13100 9016 7776 9595 ...
##  $ AreaT5       : int  10410 8105 7358 7682 8931 8518 16468 10331 7527 8454 ...
##  $ AreaT6       : int  8667 6063 6732 5718 8630 10349 11798 7986 5248 8749 ...
##  $ PercT4       : num  13.6 28.1 41.4 25.4 17.3 ...
##  $ PercT5       : num  27.8 45.8 62.7 37.4 32.8 ...
##  $ PercT6       : num  53.9 71 78 44.8 40 ...
##  $ TotalArea    : int  31516 24398 21615 28298 26721 27146 41366 27333 20551 26798 ...
##  $ TotalBlack   : num  9270 10900 12986 9220 7966 ...
##  $ TotalPerc    : num  29.4 44.7 60.1 32.6 29.8 ...
##  $ Orientation  : chr  "OK" "OK" "OK" "OK" ...

d_Pgm$Supervisor.PI <- as.factor(d_Pgm$Supervisor.PI)
d_Pgm$Diet <- as.factor(d_Pgm$Diet)
d_Pgm$Batch <- as.factor(d_Pgm$Batch)
d_Pgm$Population_Lat <- factor(d_Pgm$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_Pgm$Population_Lon <- factor(d_Pgm$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_Pgm$Population_Alt <- factor(d_Pgm$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_Pgm$Line <- as.factor(d_Pgm$Line)
d_Pgm$AreaT4 <- as.numeric(d_Pgm$AreaT4)
d_Pgm$AreaT5 <- as.numeric(d_Pgm$AreaT5)
d_Pgm$AreaT6 <- as.numeric(d_Pgm$AreaT6)
d_Pgm$PercT4 <- as.numeric(d_Pgm$PercT4)
d_Pgm$PercT5 <- as.numeric(d_Pgm$PercT5)
d_Pgm$PercT6 <- as.numeric(d_Pgm$PercT6)
d_Pgm$TotalArea <- as.numeric(d_Pgm$TotalArea)
d_Pgm$TotalBlack <- as.numeric(d_Pgm$TotalBlack)

str(d_Pgm)

## 'data.frame':    3694 obs. of  20 variables:
##  $ Supervisor.PI : Factor w/ 2 levels "Abbott","Gibert": 2 2 2 2 2 2 2 2 2 2 ...
##  $ Diet          : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch         : Factor w/ 3 levels "1","2","3": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population    : chr  "AK" "AK" "AK" "AK" ...
##  $ Line          : Factor w/ 168 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 11 ...
##  $ Sex           : logi  FALSE FALSE FALSE FALSE FALSE FALSE ...
##  $ Individual    : int  1 2 3 4 5 6 7 8 9 1 ...
##  $ AreaT4        : num  12439 10230 7525 14898 9160 ...
##  $ AreaT5        : num  10410 8105 7358 7682 8931 ...
##  $ AreaT6        : num  8667 6063 6732 5718 8630 ...
##  $ PercT4        : num  13.6 28.1 41.4 25.4 17.3 ...
##  $ PercT5        : num  27.8 45.8 62.7 37.4 32.8 ...
##  $ PercT6        : num  53.9 71 78 44.8 40 ...
##  $ TotalArea     : num  31516 24398 21615 28298 26721 ...
##  $ TotalBlack    : num  9270 10900 12986 9220 7966 ...
##  $ TotalPerc     : num  29.4 44.7 60.1 32.6 29.8 ...
##  $ Orientation   : chr  "OK" "OK" "OK" "OK" ...
##  $ Population_Lat: Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon: Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt: Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

d_Pgm2 <- read.csv("MasterSheets_Oct21_git/PGM2_MasterSheet_Oct21.csv")

str(d_Pgm2)

## 'data.frame':    1514 obs. of  15 variables:
##  $ Supervisor.PI: chr  "Schmidt" "Schmidt" "Schmidt" "Schmidt" ...
##  $ Diet         : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch        : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population   : chr  "AK" "AK" "AK" "AK" ...
##  $ Line         : chr  "AK1" "AK1" "AK1" "AK1" ...
##  $ Sex          : logi  FALSE FALSE FALSE FALSE FALSE FALSE ...
##  $ Individual   : int  2 9 1 10 3 6 8 5 4 7 ...
##  $ Tergite8     : int  4 3 2 2 2 2 2 1 1 1 ...
##  $ Tergite9     : int  8 7 6 5 4 4 4 3 2 1 ...
##  $ Tergite10    : int  6 3 3 4 1 1 1 1 0 0 ...
##  $ Total        : int  18 13 11 11 7 7 7 5 3 2 ...
##  $ Country      : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude     : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude    : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude     : int  88 88 88 88 88 88 88 88 88 88 ...

d_Pgm2$Supervisor.PI <- as.factor(d_Pgm2$Supervisor.PI)
d_Pgm2$Diet <- as.factor(d_Pgm2$Diet)
d_Pgm2$Batch <- as.factor(d_Pgm2$Batch)
d_Pgm2$Population_Lat <- factor(d_Pgm2$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_Pgm2$Population_Lon <- factor(d_Pgm2$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_Pgm2$Population_Alt <- factor(d_Pgm2$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_Pgm2$Line <- as.factor(d_Pgm2$Line)
d_Pgm2$Tergite8 <- as.numeric(d_Pgm2$Tergite8)
d_Pgm2$Tergite9 <- as.numeric(d_Pgm2$Tergite9)
d_Pgm2$Tergite10 <- as.numeric(d_Pgm2$Tergite10)
d_Pgm2$Total <- as.numeric(d_Pgm2$Total)

str(d_Pgm2)

## 'data.frame':    1514 obs. of  18 variables:
##  $ Supervisor.PI : Factor w/ 1 level "Schmidt": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet          : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch         : Factor w/ 1 level "1": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population    : chr  "AK" "AK" "AK" "AK" ...
##  $ Line          : Factor w/ 157 levels "AK1","AK10","AK11",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sex           : logi  FALSE FALSE FALSE FALSE FALSE FALSE ...
##  $ Individual    : int  2 9 1 10 3 6 8 5 4 7 ...
##  $ Tergite8      : num  4 3 2 2 2 2 2 1 1 1 ...
##  $ Tergite9      : num  8 7 6 5 4 4 4 3 2 1 ...
##  $ Tergite10     : num  6 3 3 4 1 1 1 1 0 0 ...
##  $ Total         : num  18 13 11 11 7 7 7 5 3 2 ...
##  $ Country       : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude      : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude     : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude      : int  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat: Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon: Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt: Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

Descriptive Statistics

T4

Descriptive statistics at the line level, with batch information :

table_Pgm_Line_PercT4_wbatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(PercT4), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_PercT4_wbatch.csv", row.names = T)

table_Pgm_Line_PercT4_wbatch <- read.csv("Pigmentation/table_Pgm_Line_PercT4_wbatch.csv")

DT::datatable(table_Pgm_Line_PercT4_wbatch, options = list(pageLength = 10)) %>% formatRound(6:13, 3)

Descriptive statistics at the line level, without batch information :

table_Pgm_Line_PercT4_wobatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(PercT4), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_PercT4_wobatch.csv", row.names = T)
table_Pgm_Line_PercT4_wobatch <- read.csv("Pigmentation/table_Pgm_Line_PercT4_wobatch.csv")

DT::datatable(table_Pgm_Line_PercT4_wobatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, with batch information :

table_Pgm_Pop_PercT4_wbatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Batch, Population) %>% 
                                   summarise_at(vars(PercT4), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_PercT4_wbatch.csv", row.names = T)

table_Pgm_Pop_PercT4_wbatch <- read.csv("Pigmentation/table_Pgm_Pop_PercT4_wbatch.csv")

DT::datatable(table_Pgm_Pop_PercT4_wbatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, without batch information :

table_Pgm_Pop_PercT4_wobatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Population) %>% 
                                   summarise_at(vars(PercT4), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_PercT4_wobatch.csv", row.names = T)

table_Pgm_Pop_PercT4_wobatch <- read.csv("Pigmentation/table_Pgm_Pop_PercT4_wobatch.csv")

DT::datatable(table_Pgm_Pop_PercT4_wobatch, options = list(pageLength = 10))%>% formatRound(4:11, 3)

T5

Descriptive statistics at the line level, with batch information :

table_Pgm_Line_PercT5_wbatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(PercT5), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_PercT5_wbatch.csv", row.names = T)

table_Pgm_Line_PercT5_wbatch <- read.csv("Pigmentation/table_Pgm_Line_PercT5_wbatch.csv")

DT::datatable(table_Pgm_Line_PercT5_wbatch, options = list(pageLength = 10)) %>% formatRound(6:13, 3)

Descriptive statistics at the line level, without batch information :

table_Pgm_Line_PercT5_wobatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(PercT5), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_PercT5_wobatch.csv", row.names = T)
table_Pgm_Line_PercT5_wobatch <- read.csv("Pigmentation/table_Pgm_Line_PercT5_wobatch.csv")

DT::datatable(table_Pgm_Line_PercT5_wobatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, with batch information :

table_Pgm_Pop_PercT5_wbatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Batch, Population) %>% 
                                   summarise_at(vars(PercT5), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_PercT5_wbatch.csv", row.names = T)

table_Pgm_Pop_PercT5_wbatch <- read.csv("Pigmentation/table_Pgm_Pop_PercT5_wbatch.csv")

DT::datatable(table_Pgm_Pop_PercT5_wbatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, without batch information :

table_Pgm_Pop_PercT5_wobatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Population) %>% 
                                   summarise_at(vars(PercT5), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_PercT5_wobatch.csv", row.names = T)

table_Pgm_Pop_PercT5_wobatch <- read.csv("Pigmentation/table_Pgm_Pop_PercT5_wobatch.csv")

DT::datatable(table_Pgm_Pop_PercT5_wobatch, options = list(pageLength = 10))%>% formatRound(4:11, 3)

T6

Descriptive statistics at the line level, with batch information :

table_Pgm_Line_PercT6_wbatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(PercT6), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_PercT6_wbatch.csv", row.names = T)

table_Pgm_Line_PercT6_wbatch <- read.csv("Pigmentation/table_Pgm_Line_PercT6_wbatch.csv")

DT::datatable(table_Pgm_Line_PercT6_wbatch, options = list(pageLength = 10)) %>% formatRound(6:13, 3)

Descriptive statistics at the line level, without batch information :

table_Pgm_Line_PercT6_wobatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(PercT6), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_PercT6_wobatch.csv", row.names = T)
table_Pgm_Line_PercT6_wobatch <- read.csv("Pigmentation/table_Pgm_Line_PercT6_wobatch.csv")

DT::datatable(table_Pgm_Line_PercT6_wobatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, with batch information :

table_Pgm_Pop_PercT6_wbatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Batch, Population) %>% 
                                   summarise_at(vars(PercT6), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_PercT6_wbatch.csv", row.names = T)

table_Pgm_Pop_PercT6_wbatch <- read.csv("Pigmentation/table_Pgm_Pop_PercT6_wbatch.csv")

DT::datatable(table_Pgm_Pop_PercT6_wbatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, without batch information :

table_Pgm_Pop_PercT6_wobatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Population) %>% 
                                   summarise_at(vars(PercT6), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_PercT6_wobatch.csv", row.names = T)

table_Pgm_Pop_PercT6_wobatch <- read.csv("Pigmentation/table_Pgm_Pop_PercT6_wobatch.csv")

DT::datatable(table_Pgm_Pop_PercT6_wobatch, options = list(pageLength = 10))%>% formatRound(4:11, 3)

Total

Descriptive statistics at the line level, with batch information :

table_Pgm_Line_TotalPerc_wbatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(TotalPerc), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_TotalPerc_wbatch.csv", row.names = T)

table_Pgm_Line_TotalPerc_wbatch <- read.csv("Pigmentation/table_Pgm_Line_TotalPerc_wbatch.csv")

DT::datatable(table_Pgm_Line_TotalPerc_wbatch, options = list(pageLength = 10)) %>% formatRound(6:13, 3)

Descriptive statistics at the line level, without batch information :

table_Pgm_Line_TotalPerc_wobatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(TotalPerc), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_TotalPerc_wobatch.csv", row.names = T)
table_Pgm_Line_TotalPerc_wobatch <- read.csv("Pigmentation/table_Pgm_Line_TotalPerc_wobatch.csv")

DT::datatable(table_Pgm_Line_TotalPerc_wobatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, with batch information :

table_Pgm_Pop_TotalPerc_wbatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Batch, Population) %>% 
                                   summarise_at(vars(TotalPerc), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_TotalPerc_wbatch.csv", row.names = T)

table_Pgm_Pop_TotalPerc_wbatch <- read.csv("Pigmentation/table_Pgm_Pop_TotalPerc_wbatch.csv")

DT::datatable(table_Pgm_Pop_TotalPerc_wbatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, without batch information :

table_Pgm_Pop_TotalPerc_wobatch <- write.csv(d_Pgm %>% group_by(Supervisor.PI, Population) %>% 
                                   summarise_at(vars(TotalPerc), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_TotalPerc_wobatch.csv", row.names = T)

table_Pgm_Pop_TotalPerc_wobatch <- read.csv("Pigmentation/table_Pgm_Pop_TotalPerc_wobatch.csv")

DT::datatable(table_Pgm_Pop_TotalPerc_wobatch, options = list(pageLength = 10))%>% formatRound(4:11, 3)

T8

Descriptive statistics at the line level, with batch information :

table_Pgm_Line_Tergite8_wbatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(Tergite8), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_Tergite8_wbatch.csv", row.names = T)

table_Pgm_Line_Tergite8_wbatch <- read.csv("Pigmentation/table_Pgm_Line_Tergite8_wbatch.csv")

DT::datatable(table_Pgm_Line_Tergite8_wbatch, options = list(pageLength = 10)) %>% formatRound(6:13, 3)

Descriptive statistics at the line level, without batch information :

table_Pgm_Line_Tergite8_wobatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(Tergite8), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_Tergite8_wobatch.csv", row.names = T)
table_Pgm_Line_Tergite8_wobatch <- read.csv("Pigmentation/table_Pgm_Line_Tergite8_wobatch.csv")

DT::datatable(table_Pgm_Line_Tergite8_wobatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, with batch information :

table_Pgm_Pop_Tergite8_wbatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Batch, Population) %>% 
                                   summarise_at(vars(Tergite8), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_Tergite8_wbatch.csv", row.names = T)

table_Pgm_Pop_Tergite8_wbatch <- read.csv("Pigmentation/table_Pgm_Pop_Tergite8_wbatch.csv")

DT::datatable(table_Pgm_Pop_Tergite8_wbatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, without batch information :

table_Pgm_Pop_Tergite8_wobatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Population) %>% 
                                   summarise_at(vars(Tergite8), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_Tergite8_wobatch.csv", row.names = T)

table_Pgm_Pop_Tergite8_wobatch <- read.csv("Pigmentation/table_Pgm_Pop_Tergite8_wobatch.csv")

DT::datatable(table_Pgm_Pop_Tergite8_wobatch, options = list(pageLength = 10))%>% formatRound(4:11, 3)

T9

Descriptive statistics at the line level, with batch information :

table_Pgm_Line_Tergite9_wbatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(Tergite9), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_Tergite9_wbatch.csv", row.names = T)

table_Pgm_Line_Tergite9_wbatch <- read.csv("Pigmentation/table_Pgm_Line_Tergite9_wbatch.csv")

DT::datatable(table_Pgm_Line_Tergite9_wbatch, options = list(pageLength = 10)) %>% formatRound(6:13, 3)

Descriptive statistics at the line level, without batch information :

table_Pgm_Line_Tergite9_wobatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(Tergite9), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_Tergite9_wobatch.csv", row.names = T)
table_Pgm_Line_Tergite9_wobatch <- read.csv("Pigmentation/table_Pgm_Line_Tergite9_wobatch.csv")

DT::datatable(table_Pgm_Line_Tergite9_wobatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, with batch information :

table_Pgm_Pop_Tergite9_wbatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Batch, Population) %>% 
                                   summarise_at(vars(Tergite9), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_Tergite9_wbatch.csv", row.names = T)

table_Pgm_Pop_Tergite9_wbatch <- read.csv("Pigmentation/table_Pgm_Pop_Tergite9_wbatch.csv")

DT::datatable(table_Pgm_Pop_Tergite9_wbatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, without batch information :

table_Pgm_Pop_Tergite9_wobatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Population) %>% 
                                   summarise_at(vars(Tergite9), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_Tergite9_wobatch.csv", row.names = T)

table_Pgm_Pop_Tergite9_wobatch <- read.csv("Pigmentation/table_Pgm_Pop_Tergite9_wobatch.csv")

DT::datatable(table_Pgm_Pop_Tergite9_wobatch, options = list(pageLength = 10))%>% formatRound(4:11, 3)

T10

Descriptive statistics at the line level, with batch information :

table_Pgm_Line_Tergite10_wbatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(Tergite10), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_Tergite10_wbatch.csv", row.names = T)

table_Pgm_Line_Tergite10_wbatch <- read.csv("Pigmentation/table_Pgm_Line_Tergite10_wbatch.csv")

DT::datatable(table_Pgm_Line_Tergite10_wbatch, options = list(pageLength = 10)) %>% formatRound(6:13, 3)

Descriptive statistics at the line level, without batch information :

table_Pgm_Line_Tergite10_wobatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(Tergite10), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_Tergite10_wobatch.csv", row.names = T)
table_Pgm_Line_Tergite10_wobatch <- read.csv("Pigmentation/table_Pgm_Line_Tergite10_wobatch.csv")

DT::datatable(table_Pgm_Line_Tergite10_wobatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, with batch information :

table_Pgm_Pop_Tergite10_wbatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Batch, Population) %>% 
                                   summarise_at(vars(Tergite10), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_Tergite10_wbatch.csv", row.names = T)

table_Pgm_Pop_Tergite10_wbatch <- read.csv("Pigmentation/table_Pgm_Pop_Tergite10_wbatch.csv")

DT::datatable(table_Pgm_Pop_Tergite10_wbatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, without batch information :

table_Pgm_Pop_Tergite10_wobatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Population) %>% 
                                   summarise_at(vars(Tergite10), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_Tergite10_wobatch.csv", row.names = T)

table_Pgm_Pop_Tergite10_wobatch <- read.csv("Pigmentation/table_Pgm_Pop_Tergite10_wobatch.csv")

DT::datatable(table_Pgm_Pop_Tergite10_wobatch, options = list(pageLength = 10))%>% formatRound(4:11, 3)

Total2

Descriptive statistics at the line level, with batch information :

table_Pgm_Line_Total_wbatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Batch, Population, Line) %>% summarise_at(vars(Total), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_Total_wbatch.csv", row.names = T)

table_Pgm_Line_Total_wbatch <- read.csv("Pigmentation/table_Pgm_Line_Total_wbatch.csv")

DT::datatable(table_Pgm_Line_Total_wbatch, options = list(pageLength = 10)) %>% formatRound(6:13, 3)

Descriptive statistics at the line level, without batch information :

table_Pgm_Line_Total_wobatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Population, Line) %>% summarise_at(vars(Total), list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, SE = std_err, CV = coef_var, Mode = estimate_mode)), file = "Pigmentation/table_Pgm_Line_Total_wobatch.csv", row.names = T)
table_Pgm_Line_Total_wobatch <- read.csv("Pigmentation/table_Pgm_Line_Total_wobatch.csv")

DT::datatable(table_Pgm_Line_Total_wobatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, with batch information :

table_Pgm_Pop_Total_wbatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Batch, Population) %>% 
                                   summarise_at(vars(Total), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_Total_wbatch.csv", row.names = T)

table_Pgm_Pop_Total_wbatch <- read.csv("Pigmentation/table_Pgm_Pop_Total_wbatch.csv")

DT::datatable(table_Pgm_Pop_Total_wbatch, options = list(pageLength = 10)) %>% formatRound(5:12, 3)

Descriptive statistics at the population level, without batch information :

table_Pgm_Pop_Total_wobatch <- write.csv(d_Pgm2 %>% group_by(Supervisor.PI, Population) %>% 
                                   summarise_at(vars(Total), 
                                                list(Mean = mean, SD= sd, Median = median, Min = min, Max = max, 
                                                     SE = std_err, CV = coef_var, Mode = estimate_mode)), 
                           file = "Pigmentation/table_Pgm_Pop_Total_wobatch.csv", row.names = T)

table_Pgm_Pop_Total_wobatch <- read.csv("Pigmentation/table_Pgm_Pop_Total_wobatch.csv")

DT::datatable(table_Pgm_Pop_Total_wobatch, options = list(pageLength = 10))%>% formatRound(4:11, 3)

Overview

Plots and Linear Models by Lab

Abbott Lab

PercT4

Pgm_lmer_Abbott_PercT4 <- lmer(PercT4 ~ -1 + Population + (1|Line:Population), 
                               data = (subset(d_Pgm,Supervisor.PI=='Abbott')))
capture.output(summary(Pgm_lmer_Abbott_PercT4),file = "Pigmentation/Pgm_lmer_Abbott_PercT4_sum.txt")
capture.output(anova(Pgm_lmer_Abbott_PercT4),file = "Pigmentation/Pgm_lmer_Abbott_PercT4.txt")

anova(Pgm_lmer_Abbott_PercT4)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population  57862  6429.1     9 151.93  208.03 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm_lmer_Abbott_PercT4)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: PercT4 ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_Pgm, Supervisor.PI == "Abbott"))
## 
## REML criterion at convergence: 12537.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.1997 -0.5926 -0.1274  0.4734  7.9312 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 16.14    4.018   
##  Residual                    30.91    5.559   
## Number of obs: 1953, groups:  Line:Population, 161
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  14.2768     0.9638 149.7585  14.813  < 2e-16 ***
## PopulationGI  13.2016     1.1577 152.7995  11.403  < 2e-16 ***
## PopulationKA  10.1518     0.9932 152.5033  10.221  < 2e-16 ***
## PopulationMA  21.6093     1.0211 152.8771  21.162  < 2e-16 ***
## PopulationMU  15.2126     1.0484 151.5568  14.510  < 2e-16 ***
## PopulationRE  17.3464     1.0829 152.7637  16.018  < 2e-16 ***
## PopulationUM  14.6774     1.0192 151.7714  14.400  < 2e-16 ***
## PopulationVA   7.5382     0.9901 150.5722   7.614 2.72e-12 ***
## PopulationYE  14.9440     0.9687 152.8572  15.428  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Pgm_Abbott_PercT4 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Abbott')), aes(x=Population_Lat, y=PercT4, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_Abbott_PercT4", x="Population", y = "PercT4") + ylim(c(min(d_Pgm$PercT4) , max(d_Pgm$PercT4)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm_pop_Abbott_PercT4 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Abbott')), aes(x=Population_Lat, y=PercT4, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_pop_Abbott_PercT4", x="Population", y = "PercT4") + ylim(c(min(d_Pgm$PercT4) , max(d_Pgm$PercT4)))+ droseu_theme

## quartz_off_screen 
##                 2

PercT5

Pgm_lmer_Abbott_PercT5 <- lmer(PercT5 ~ -1 + Population  + (1|Line:Population) , data = (subset(d_Pgm,Supervisor.PI=='Abbott')))
capture.output(summary(Pgm_lmer_Abbott_PercT5),file = "Pigmentation/Pgm_lmer_Abbott_PercT5_sum.txt")
capture.output(anova(Pgm_lmer_Abbott_PercT5),file = "Pigmentation/Pgm_lmer_Abbott_PercT5.txt")

anova(Pgm_lmer_Abbott_PercT5)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population  88573  9841.4     9 151.24  284.41 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm_lmer_Abbott_PercT5)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: PercT5 ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_Pgm, Supervisor.PI == "Abbott"))
## 
## REML criterion at convergence: 12736.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.1419 -0.6487 -0.1115  0.5154  6.4092 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 15.44    3.929   
##  Residual                    34.60    5.882   
## Number of obs: 1953, groups:  Line:Population, 161
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK  17.7494     0.9531 148.7407   18.62   <2e-16 ***
## PopulationGI  15.8249     1.1456 152.2199   13.81   <2e-16 ***
## PopulationKA  12.6107     0.9828 151.8949   12.83   <2e-16 ***
## PopulationMA  23.6705     1.0105 152.3113   23.43   <2e-16 ***
## PopulationMU  17.3056     1.0372 150.8151   16.68   <2e-16 ***
## PopulationRE  18.4126     1.0716 152.1776   17.18   <2e-16 ***
## PopulationUM  17.7191     1.0084 151.0598   17.57   <2e-16 ***
## PopulationVA  10.4377     0.9792 149.6941   10.66   <2e-16 ***
## PopulationYE  16.9507     0.9586 152.2972   17.68   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Pgm_Abbott_PercT5 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Abbott')), aes(x=Population_Lat, y=PercT5, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_Abbott_PercT5", x="Population", y = "PercT5") + ylim(c(min(d_Pgm$PercT5) , max(d_Pgm$PercT5)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm_pop_Abbott_PercT5 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Abbott')), aes(x=Population_Lat, y=PercT5, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_pop_Abbott_PercT5", x="Population", y = "PercT5") + ylim(c(min(d_Pgm$PercT5) , max(d_Pgm$PercT5)))+ droseu_theme

## quartz_off_screen 
##                 2

PercT6

Pgm_lmer_Abbott_PercT6 <- lmer(PercT6 ~ -1 + Population  + (1|Line:Population) , data = (subset(d_Pgm,Supervisor.PI=='Abbott')))
capture.output(summary(Pgm_lmer_Abbott_PercT6),file = "Pigmentation/Pgm_lmer_Abbott_PercT6_sum.txt")
capture.output(anova(Pgm_lmer_Abbott_PercT6),file = "Pigmentation/Pgm_lmer_Abbott_PercT6.txt")

anova(Pgm_lmer_Abbott_PercT6)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 246476   27386     9 151.69  171.06 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm_lmer_Abbott_PercT6)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: PercT6 ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_Pgm, Supervisor.PI == "Abbott"))
## 
## REML criterion at convergence: 15794.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.3191 -0.6368 -0.0337  0.5949  3.1370 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 130.3    11.42   
##  Residual                    160.1    12.65   
## Number of obs: 1953, groups:  Line:Population, 161
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   34.383      2.674 150.252   12.86   <2e-16 ***
## PopulationGI   33.934      3.206 152.282   10.58   <2e-16 ***
## PopulationKA   32.871      2.751 152.066   11.95   <2e-16 ***
## PopulationMA   47.543      2.828 152.330   16.81   <2e-16 ***
## PopulationMU   31.476      2.906 151.430   10.83   <2e-16 ***
## PopulationRE   42.547      2.999 152.260   14.19   <2e-16 ***
## PopulationUM   41.836      2.824 151.574   14.81   <2e-16 ***
## PopulationVA   33.891      2.745 150.766   12.34   <2e-16 ***
## PopulationYE   32.465      2.683 152.305   12.10   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Pgm_Abbott_PercT6 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Abbott')), aes(x=Population_Lat, y=PercT6, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_Abbott_PercT6", x="Population", y = "PercT6") + ylim(c(min(d_Pgm$PercT6) , max(d_Pgm$PercT6)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm_pop_Abbott_PercT6 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Abbott')), aes(x=Population_Lat, y=PercT6, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_pop_Abbott_PercT6", x="Population", y = "PercT6") + ylim(c(min(d_Pgm$PercT6) , max(d_Pgm$PercT6)))+ droseu_theme

## quartz_off_screen 
##                 2

TotalPerc

Pgm_lmer_Abbott_TotalPerc <- lmer(TotalPerc ~ -1 + Population  + (1|Line:Population)  , data = (subset(d_Pgm,Supervisor.PI=='Abbott')))
capture.output(summary(Pgm_lmer_Abbott_TotalPerc),file = "Pigmentation/Pgm_lmer_Abbott_TotalPerc_sum.txt")
capture.output(anova(Pgm_lmer_Abbott_TotalPerc),file = "Pigmentation/Pgm_lmer_Abbott_TotalPerc.txt")

anova(Pgm_lmer_Abbott_TotalPerc)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 100129   11126     9 151.48  325.16 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm_lmer_Abbott_TotalPerc)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TotalPerc ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_Pgm, Supervisor.PI == "Abbott"))
## 
## REML criterion at convergence: 12772.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.7535 -0.6401 -0.0873  0.5628  4.8794 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 23.64    4.862   
##  Residual                    34.22    5.849   
## Number of obs: 1953, groups:  Line:Population, 161
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   21.019      1.148 149.802   18.32   <2e-16 ***
## PopulationGI   19.740      1.377 152.160   14.34   <2e-16 ***
## PopulationKA   17.614      1.181 151.915   14.91   <2e-16 ***
## PopulationMA   30.027      1.214 152.217   24.73   <2e-16 ***
## PopulationMU   20.547      1.247 151.179   16.47   <2e-16 ***
## PopulationRE   25.145      1.288 152.133   19.52   <2e-16 ***
## PopulationUM   23.795      1.213 151.346   19.62   <2e-16 ***
## PopulationVA   16.142      1.178 150.410   13.70   <2e-16 ***
## PopulationYE   20.815      1.152 152.192   18.07   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Pgm_Abbott_TotalPerc <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Abbott')), aes(x=Population_Lat, y=TotalPerc, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_Abbott_TotalPerc", x="Population", y = "TotalPerc") + ylim(c(min(d_Pgm$TotalPerc) , max(d_Pgm$TotalPerc)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm_pop_Abbott_TotalPerc <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Abbott')), aes(x=Population_Lat, y=TotalPerc, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_pop_Abbott_TotalPerc", x="Population", y = "TotalPerc") + ylim(c(min(d_Pgm$TotalPerc) , max(d_Pgm$TotalPerc)))+ droseu_theme

## quartz_off_screen 
##                 2

Gibert Lab

PercT4

Pgm_lmer_Gibert_PercT4 <- lmer(PercT4 ~ -1 + Population  + (1|Line:Population) +
                                    (1|Batch), data = (subset(d_Pgm,Supervisor.PI=='Gibert')))
capture.output(summary(Pgm_lmer_Gibert_PercT4),file = "Pigmentation/Pgm_lmer_Gibert_PercT4_sum.txt")
capture.output(anova(Pgm_lmer_Gibert_PercT4),file = "Pigmentation/Pgm_lmer_Gibert_PercT4.txt")

anova(Pgm_lmer_Gibert_PercT4)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value  Pr(>F)  
## Population  33916  3768.4     9     2  41.486 0.02375 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm_lmer_Gibert_PercT4)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: PercT4 ~ -1 + Population + (1 | Line:Population) + (1 | Batch)
##    Data: (subset(d_Pgm, Supervisor.PI == "Gibert"))
## 
## REML criterion at convergence: 12935.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.4920 -0.6719 -0.1283  0.5731  5.9803 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 15.601   3.950   
##  Batch           (Intercept)  3.226   1.796   
##  Residual                    90.836   9.531   
## Number of obs: 1741, groups:  Line:Population, 167; Batch, 3
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   24.283      1.579  4.554   15.37 4.25e-05 ***
## PopulationGI   23.395      1.761  5.580   13.28 1.96e-05 ***
## PopulationKA   20.859      1.641  4.220   12.71  0.00016 ***
## PopulationMA   23.780      1.665  4.464   14.29 6.76e-05 ***
## PopulationMU   25.072      1.642  4.227   15.27 7.39e-05 ***
## PopulationRE   19.969      1.731  5.218   11.53 6.54e-05 ***
## PopulationUM   24.473      1.681  4.647   14.55 4.70e-05 ***
## PopulationVA   24.307      1.661  4.422   14.64 6.48e-05 ***
## PopulationYE   25.401      1.654  4.350   15.36 5.91e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.487                                                  
## PopulatinKA 0.522  0.509                                           
## PopulatinMA 0.515  0.501  0.539                                    
## PopulatinMU 0.522  0.509  0.547  0.539                             
## PopulatinRE 0.496  0.485  0.518  0.510  0.518                      
## PopulatinUM 0.510  0.497  0.534  0.526  0.534  0.505               
## PopulatinVA 0.516  0.502  0.541  0.533  0.540  0.511  0.528        
## PopulatinYE 0.518  0.505  0.543  0.535  0.542  0.514  0.530  0.536

p_Pgm_Gibert_PercT4 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=PercT4, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_Gibert_PercT4", x="Population", y = "PercT4") + ylim(c(min(d_Pgm$PercT4) , max(d_Pgm$PercT4)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm_pop_Gibert_PercT4 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=PercT4, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_pop_Gibert_PercT4", x="Population", y = "PercT4") + ylim(c(min(d_Pgm$PercT4) , max(d_Pgm$PercT4)))+ droseu_theme

## quartz_off_screen 
##                 2

PercT5

Pgm_lmer_Gibert_PercT5 <- lmer(PercT5 ~ -1 + Population  + (1|Line:Population) +
                                    (1|Batch), data = (subset(d_Pgm,Supervisor.PI=='Gibert')))

## boundary (singular) fit: see ?isSingular

capture.output(summary(Pgm_lmer_Gibert_PercT5),file = "Pigmentation/Pgm_lmer_Gibert_PercT5_sum.txt")
capture.output(anova(Pgm_lmer_Gibert_PercT5),file = "Pigmentation/Pgm_lmer_Gibert_PercT5.txt")

anova(Pgm_lmer_Gibert_PercT5)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 399387   44376     9 157.33  420.43 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm_lmer_Gibert_PercT5)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: PercT5 ~ -1 + Population + (1 | Line:Population) + (1 | Batch)
##    Data: (subset(d_Pgm, Supervisor.PI == "Gibert"))
## 
## REML criterion at convergence: 13270.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.9328 -0.6447 -0.0768  0.5422  5.7227 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept)  36.17    6.014  
##  Batch           (Intercept)   0.00    0.000  
##  Residual                    105.55   10.274  
## Number of obs: 1741, groups:  Line:Population, 167; Batch, 3
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   32.603      1.525 156.954   21.39   <2e-16 ***
## PopulationGI   29.152      1.757 155.855   16.59   <2e-16 ***
## PopulationKA   29.325      1.524 156.586   19.25   <2e-16 ***
## PopulationMA   36.393      1.568 158.331   23.20   <2e-16 ***
## PopulationMU   34.539      1.525 156.954   22.66   <2e-16 ***
## PopulationRE   30.902      1.702 156.037   18.16   <2e-16 ***
## PopulationUM   32.690      1.605 156.342   20.36   <2e-16 ***
## PopulationVA   32.058      1.563 156.470   20.51   <2e-16 ***
## PopulationYE   33.204      1.541 162.629   21.55   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

p_Pgm_Gibert_PercT5 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=PercT5, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_Gibert_PercT5", x="Population", y = "PercT5") + ylim(c(min(d_Pgm$PercT5) , max(d_Pgm$PercT5)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm_pop_Gibert_PercT5 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=PercT5, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_pop_Gibert_PercT5", x="Population", y = "PercT5") + ylim(c(min(d_Pgm$PercT5) , max(d_Pgm$PercT5)))+ droseu_theme

## quartz_off_screen 
##                 2

PercT6

Pgm_lmer_Gibert_PercT6 <- lmer(PercT6 ~ -1 + Population  + (1|Line:Population) +
                                    (1|Batch), data = (subset(d_Pgm,Supervisor.PI=='Gibert')))
capture.output(summary(Pgm_lmer_Gibert_PercT6),file = "Pigmentation/Pgm_lmer_Gibert_PercT6_sum.txt")
capture.output(anova(Pgm_lmer_Gibert_PercT6),file = "Pigmentation/Pgm_lmer_Gibert_PercT6.txt")

anova(Pgm_lmer_Gibert_PercT6)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value  Pr(>F)  
## Population 100142   11127     9     2  53.671 0.01842 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm_lmer_Gibert_PercT6)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: PercT6 ~ -1 + Population + (1 | Line:Population) + (1 | Batch)
##    Data: (subset(d_Pgm, Supervisor.PI == "Gibert"))
## 
## REML criterion at convergence: 14560.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.0930 -0.6255  0.0378  0.6407  2.8805 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 173.498  13.172  
##  Batch           (Intercept)   9.633   3.104  
##  Residual                    207.318  14.399  
## Number of obs: 1741, groups:  Line:Population, 167; Batch, 3
## 
## Fixed effects:
##              Estimate Std. Error     df t value Pr(>|t|)    
## PopulationAK   46.726      3.687 10.340   12.67 1.25e-07 ***
## PopulationGI   46.167      4.191 12.542   11.02 8.20e-08 ***
## PopulationKA   58.051      3.783  8.488   15.34 1.74e-07 ***
## PopulationMA   55.051      3.855  9.126   14.28 1.49e-07 ***
## PopulationMU   52.365      3.784  8.495   13.84 4.03e-07 ***
## PopulationRE   54.235      4.092 11.460   13.26 2.65e-08 ***
## PopulationUM   53.626      3.922  9.756   13.67 1.10e-07 ***
## PopulationVA   51.340      3.850  9.082   13.33 2.86e-07 ***
## PopulationYE   49.100      3.798  8.622   12.93 6.11e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.272                                                  
## PopulatinKA 0.301  0.291                                           
## PopulatinMA 0.295  0.285  0.317                                    
## PopulatinMU 0.301  0.291  0.323  0.317                             
## PopulatinRE 0.279  0.271  0.299  0.293  0.299                      
## PopulatinUM 0.290  0.281  0.312  0.306  0.312  0.288               
## PopulatinVA 0.296  0.286  0.318  0.312  0.318  0.293  0.307        
## PopulatinYE 0.300  0.290  0.322  0.316  0.322  0.297  0.311  0.317

p_Pgm_Gibert_PercT6 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=PercT6, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_Gibert_PercT6", x="Population", y = "PercT6") + ylim(c(min(d_Pgm$PercT6) , max(d_Pgm$PercT6)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm_pop_Gibert_PercT6 <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=PercT6, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_pop_Gibert_PercT6", x="Population", y = "PercT6") + ylim(c(min(d_Pgm$PercT6) , max(d_Pgm$PercT6)))+ droseu_theme

## quartz_off_screen 
##                 2

TotalPerc

Pgm_lmer_Gibert_TotalPerc <- lmer(TotalPerc ~ -1 + Population  + (1|Line:Population) +
                                    (1|Batch), data = (subset(d_Pgm,Supervisor.PI=='Gibert')))

## boundary (singular) fit: see ?isSingular

capture.output(summary(Pgm_lmer_Gibert_TotalPerc),file = "Pigmentation/Pgm_lmer_Gibert_TotalPerc_sum.txt")
capture.output(anova(Pgm_lmer_Gibert_TotalPerc),file = "Pigmentation/Pgm_lmer_Gibert_TotalPerc.txt")

anova(Pgm_lmer_Gibert_TotalPerc)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 348622   38736     9 157.95   526.8 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm_lmer_Gibert_TotalPerc)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: TotalPerc ~ -1 + Population + (1 | Line:Population) + (1 | Batch)
##    Data: (subset(d_Pgm, Supervisor.PI == "Gibert"))
## 
## REML criterion at convergence: 12682.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.2632 -0.6733 -0.0496  0.6084  4.5241 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 34.06    5.836   
##  Batch           (Intercept)  0.00    0.000   
##  Residual                    73.53    8.575   
## Number of obs: 1741, groups:  Line:Population, 167; Batch, 3
## 
## Fixed effects:
##              Estimate Std. Error      df t value Pr(>|t|)    
## PopulationAK   33.307      1.436 157.600   23.19   <2e-16 ***
## PopulationGI   31.425      1.656 156.755   18.97   <2e-16 ***
## PopulationKA   33.445      1.435 157.305   23.30   <2e-16 ***
## PopulationMA   36.135      1.477 158.778   24.47   <2e-16 ***
## PopulationMU   35.380      1.436 157.600   24.64   <2e-16 ***
## PopulationRE   33.354      1.604 156.892   20.80   <2e-16 ***
## PopulationUM   35.303      1.513 157.122   23.34   <2e-16 ***
## PopulationVA   34.543      1.473 157.218   23.46   <2e-16 ***
## PopulationYE   34.505      1.449 162.399   23.82   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000 
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

p_Pgm_Gibert_TotalPerc <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=TotalPerc, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_Gibert_TotalPerc", x="Population", y = "TotalPerc") + ylim(c(min(d_Pgm$TotalPerc) , max(d_Pgm$TotalPerc)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm_pop_Gibert_TotalPerc <- ggplot(data = (subset(d_Pgm,Supervisor.PI=='Gibert')), aes(x=Population_Lat, y=TotalPerc, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm_pop_Gibert_TotalPerc", x="Population", y = "TotalPerc") + ylim(c(min(d_Pgm$TotalPerc) , max(d_Pgm$TotalPerc)))+ droseu_theme

## quartz_off_screen 
##                 2

Schmidt Lab

Tergite8

Pgm2_lmer_Schmidt_Tergite8 <- lmer(Tergite8 ~ -1 + Population  + (1|Line:Population), data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')))
capture.output(summary(Pgm2_lmer_Schmidt_Tergite8),file = "Pigmentation/Pgm2_lmer_Schmidt_Tergite8_sum.txt")
capture.output(anova(Pgm2_lmer_Schmidt_Tergite8),file = "Pigmentation/Pgm2_lmer_Schmidt_Tergite8.txt")

anova(Pgm2_lmer_Schmidt_Tergite8)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 1069.9  118.88     9 147.4  165.38 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm2_lmer_Schmidt_Tergite8)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: Tergite8 ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_Pgm2, Supervisor.PI == "Schmidt"))
## 
## REML criterion at convergence: 4128.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.3233 -0.6262 -0.0680  0.5288  7.2893 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.5292   0.7274  
##  Residual                    0.7188   0.8478  
## Number of obs: 1514, groups:  Line:Population, 157
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK   2.2700     0.1734 145.4931  13.094  < 2e-16 ***
## PopulationGI   2.0095     0.2165 149.3116   9.282  < 2e-16 ***
## PopulationKA   2.4947     0.1779 145.4931  14.026  < 2e-16 ***
## PopulationMA   2.8561     0.1786 147.7121  15.988  < 2e-16 ***
## PopulationMU   2.2815     0.1736 146.1725  13.144  < 2e-16 ***
## PopulationRE   2.4794     0.2091 150.6759  11.855  < 2e-16 ***
## PopulationUM   2.8021     0.2007 146.8367  13.963  < 2e-16 ***
## PopulationVA   1.6406     0.1882 145.9781   8.717 5.78e-15 ***
## PopulationYE   2.4311     0.1745 149.1257  13.935  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Pgm2_Schmidt_Tergite8 <- ggplot(data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=Tergite8, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm2_Schmidt_Tergite8", x="Population", y = "Tergite8") + ylim(c(min(d_Pgm2$Tergite8) , max(d_Pgm2$Tergite8)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm2_pop_Schmidt_Tergite8 <- ggplot(data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=Tergite8, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm2_pop_Schmidt_Tergite8", x="Population", y = "Tergite8") + ylim(c(min(d_Pgm2$Tergite8) , max(d_Pgm2$Tergite8)))+ droseu_theme

## quartz_off_screen 
##                 2

Tergite9

Pgm2_lmer_Schmidt_Tergite9 <- lmer(Tergite9 ~ -1 + Population  + (1|Line:Population) , data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')))
capture.output(summary(Pgm2_lmer_Schmidt_Tergite9),file = "Pigmentation/Pgm2_lmer_Schmidt_Tergite9_sum.txt")
capture.output(anova(Pgm2_lmer_Schmidt_Tergite9),file = "Pigmentation/Pgm2_lmer_Schmidt_Tergite9.txt")

anova(Pgm2_lmer_Schmidt_Tergite9)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 2617.1  290.79     9 148.09  114.11 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm2_lmer_Schmidt_Tergite9)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: Tergite9 ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_Pgm2, Supervisor.PI == "Schmidt"))
## 
## REML criterion at convergence: 6123.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.2209 -0.6482 -0.0807  0.5799  3.6192 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 3.671    1.916   
##  Residual                    2.548    1.596   
## Number of obs: 1514, groups:  Line:Population, 157
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK   4.5450     0.4430 146.9794  10.258  < 2e-16 ***
## PopulationGI   4.1449     0.5516 149.1735   7.514 4.91e-12 ***
## PopulationKA   5.0316     0.4546 146.9794  11.069  < 2e-16 ***
## PopulationMA   5.5072     0.4557 148.3570  12.085  < 2e-16 ***
## PopulationMU   4.8823     0.4433 147.3678  11.012  < 2e-16 ***
## PopulationRE   5.5056     0.5323 149.9741  10.343  < 2e-16 ***
## PopulationUM   6.3034     0.5123 147.7640  12.304  < 2e-16 ***
## PopulationVA   4.8631     0.4808 147.2518  10.115  < 2e-16 ***
## PopulationYE   4.7625     0.4446 149.0248  10.712  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Pgm2_Schmidt_Tergite9 <- ggplot(data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=Tergite9, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm2_Schmidt_Tergite9", x="Population", y = "Tergite9") + ylim(c(min(d_Pgm2$Tergite9) , max(d_Pgm2$Tergite9)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm2_pop_Schmidt_Tergite9 <- ggplot(data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=Tergite9, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm2_pop_Schmidt_Tergite9", x="Population", y = "Tergite9") + ylim(c(min(d_Pgm2$Tergite9) , max(d_Pgm2$Tergite9)))+ droseu_theme

## quartz_off_screen 
##                 2

Tergite10

Pgm2_lmer_Schmidt_Tergite10 <- lmer(Tergite10 ~ -1 + Population  + (1|Line:Population), data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')))
capture.output(summary(Pgm2_lmer_Schmidt_Tergite10),file = "Pigmentation/Pgm2_lmer_Schmidt_Tergite10_sum.txt")
capture.output(anova(Pgm2_lmer_Schmidt_Tergite10),file = "Pigmentation/Pgm2_lmer_Schmidt_Tergite10.txt")

anova(Pgm2_lmer_Schmidt_Tergite10)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population   1634  181.56     9 148.34   43.01 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm2_lmer_Schmidt_Tergite10)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: Tergite10 ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_Pgm2, Supervisor.PI == "Schmidt"))
## 
## REML criterion at convergence: 6832.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.7236 -0.6439 -0.1671  0.6539  3.2399 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 4.184    2.045   
##  Residual                    4.221    2.055   
## Number of obs: 1514, groups:  Line:Population, 157
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK   2.1450     0.4799 146.8229   4.470 1.56e-05 ***
## PopulationGI   2.2609     0.5984 149.8458   3.778 0.000227 ***
## PopulationKA   3.5737     0.4924 146.8229   7.258 2.11e-11 ***
## PopulationMA   3.7757     0.4941 148.6485   7.642 2.43e-12 ***
## PopulationMU   2.9894     0.4803 147.3594   6.223 4.79e-09 ***
## PopulationRE   3.0707     0.5778 150.9371   5.315 3.77e-07 ***
## PopulationUM   4.7811     0.5552 147.8953   8.611 9.93e-15 ***
## PopulationVA   3.7539     0.5209 147.2024   7.207 2.76e-11 ***
## PopulationYE   3.3644     0.4823 149.6693   6.976 9.12e-11 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Pgm2_Schmidt_Tergite10 <- ggplot(data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=Tergite10, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm2_Schmidt_Tergite10", x="Population", y = "Tergite10") + ylim(c(min(d_Pgm2$Tergite10) , max(d_Pgm2$Tergite10)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm2_pop_Schmidt_Tergite10 <- ggplot(data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=Tergite10, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm2_pop_Schmidt_Tergite10", x="Population", y = "Tergite10") + ylim(c(min(d_Pgm2$Tergite10) , max(d_Pgm2$Tergite10)))+ droseu_theme

## quartz_off_screen 
##                 2

Total

Pgm2_lmer_Schmidt_Total <- lmer(Total ~ -1 + Population  + (1|Line:Population), data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')))
capture.output(summary(Pgm2_lmer_Schmidt_Total),file = "Pigmentation/Pgm2_lmer_Schmidt_Total_sum.txt")
capture.output(anova(Pgm2_lmer_Schmidt_Total),file = "Pigmentation/Pgm2_lmer_Schmidt_Total.txt")

anova(Pgm2_lmer_Schmidt_Total)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population  12734  1414.9     9 148.07  104.26 < 2.2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Pgm2_lmer_Schmidt_Total)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: Total ~ -1 + Population + (1 | Line:Population)
##    Data: (subset(d_Pgm2, Supervisor.PI == "Schmidt"))
## 
## REML criterion at convergence: 8630.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.7804 -0.6403 -0.1225  0.6108  4.0733 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 18.12    4.257   
##  Residual                    13.57    3.684   
## Number of obs: 1514, groups:  Line:Population, 157
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## PopulationAK   8.9600     0.9869 146.8837   9.079 6.73e-16 ***
## PopulationGI   8.4120     1.2291 149.2273   6.844 1.87e-10 ***
## PopulationKA  11.1000     1.0126 146.8837  10.962  < 2e-16 ***
## PopulationMA  12.1458     1.0153 148.3448  11.963  < 2e-16 ***
## PopulationMU  10.1531     0.9876 147.2988  10.280  < 2e-16 ***
## PopulationRE  11.0455     1.1862 150.0807   9.312  < 2e-16 ***
## PopulationUM  13.8863     1.1413 147.7205  12.167  < 2e-16 ***
## PopulationVA  10.2588     1.0710 147.1752   9.579  < 2e-16 ***
## PopulationYE  10.5582     0.9907 149.0724  10.658  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             PpltAK PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI 0.000                                                  
## PopulatinKA 0.000  0.000                                           
## PopulatinMA 0.000  0.000  0.000                                    
## PopulatinMU 0.000  0.000  0.000  0.000                             
## PopulatinRE 0.000  0.000  0.000  0.000  0.000                      
## PopulatinUM 0.000  0.000  0.000  0.000  0.000  0.000               
## PopulatinVA 0.000  0.000  0.000  0.000  0.000  0.000  0.000        
## PopulatinYE 0.000  0.000  0.000  0.000  0.000  0.000  0.000  0.000

p_Pgm2_Schmidt_Total <- ggplot(data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=Total, fill=Batch)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm2_Schmidt_Total", x="Population", y = "Total") + ylim(c(min(d_Pgm2$Total) , max(d_Pgm2$Total)))+ droseu_theme

## quartz_off_screen 
##                 2

p_Pgm2_pop_Schmidt_Total <- ggplot(data = (subset(d_Pgm2,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=Total, fill=Population)) + 
        geom_boxplot(outlier.shape = NA, notch=FALSE) + 
        labs(title="p_Pgm2_pop_Schmidt_Total", x="Population", y = "Total") + ylim(c(min(d_Pgm2$Total) , max(d_Pgm2$Total)))+ droseu_theme

## quartz_off_screen 
##                 2

Lab correlations

intersect_Line_Pgm  <- Reduce(intersect, list(d_Pgm$Line[d_Pgm$Supervisor.PI == "Abbott"],
                                         d_Pgm2$Line[d_Pgm2$Supervisor.PI == "Schmidt"],
                                         d_Pgm$Line[d_Pgm$Supervisor.PI == "Gibert"]))

write(intersect_Line_Pgm, file ="Pigmentation/intersect_Line_Pgm.txt", sep = "\t")


intersect_Pop_Pgm  <- Reduce(intersect, list(d_Pgm$Population[d_Pgm$Supervisor.PI == "Abbott"],
                                         d_Pgm2$Population[d_Pgm2$Supervisor.PI == "Schmidt"],
                                         d_Pgm$Population[d_Pgm$Supervisor.PI == "Gibert"]))

write(intersect_Pop_Pgm, file ="Pigmentation/intersect_Pop_Pgm", sep = "\t")

There are 9 populations and 149 isofemale lines that have been phenotyped across the labs. Please refer to tables named intersect_Trait_level (i.e. intersect_Via_pop) for details.

Pgm_TotalPerc_fitted <- combineFitted(labs = c("Abbott", "Gibert", "Schmidt"), sex = rep("F", 3), models = list(Pgm_lmer_Abbott_TotalPerc, Pgm_lmer_Gibert_TotalPerc, Pgm2_lmer_Schmidt_Total))
  
p_Pgm_TotalPerc_lab_cor <- function() scatterPlotMatrix(Pgm_TotalPerc_fitted, sex = "F")

p_Pgm_TotalPerc_lab_cor()

pdf(file="Pigmentation/p_Pgm_TotalPerc_lab_cor.pdf", width=8, height=8)
p_Pgm_TotalPerc_lab_cor()
invisible(dev.off())

Results II : Meta-Analyses

with Diet

with Wolbachia

Viability

Via_effects <- Via_fitted %>%
  dplyr::select(-Value) %>%
  mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")), Lab = as.factor(Lab), V = SE^2, Study = paste(Population, Lab, sep = "_")) %>%
  relocate(Population, Sex, Lab, Study) %>%
  arrange(Population) %>% 
  dplyr::rename(Y = Estimate)

# run the analysis with metagen package
Via_metagen <- metagen(data = Via_effects %>% filter(Sex == "NA"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
Via_metagen <- update.meta(Via_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(Via_metagen), file = "Viability/Via_metagen_sum.txt")
Via_metagen

## Number of studies combined: k = 48
## 
##                         SMD           95%-CI     z p-value
## Random effects model 0.6535 [0.6319; 0.6752] 59.03       0
## 
## Quantifying heterogeneity:
##  tau^2 = 0.0047 [0.0030; 0.0081]; tau = 0.0686 [0.0544; 0.0902]
##  I^2 = 82.5% [77.4%; 86.4%]; H = 2.39 [2.10; 2.71]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  268.03   47 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k    SMD           95%-CI  tau^2    tau     Q   I^2
## Population = YE   5 0.5266 [0.4994; 0.5539]      0      0  2.81  0.0%
## Population = RE   6 0.6358 [0.6097; 0.6619]      0      0  0.48  0.0%
## Population = GI   5 0.6010 [0.5618; 0.6402] 0.0006 0.0254  5.94 32.7%
## Population = MU   6 0.6817 [0.6338; 0.7296] 0.0026 0.0506 18.92 73.6%
## Population = MA   5 0.6539 [0.5965; 0.7113] 0.0032 0.0562 16.57 75.9%
## Population = UM   5 0.6957 [0.6336; 0.7577] 0.0037 0.0608 17.02 76.5%
## Population = KA   5 0.7199 [0.6830; 0.7568] 0.0007 0.0273  6.99 42.8%
## Population = VA   5 0.6429 [0.5836; 0.7022] 0.0035 0.0588 17.85 77.6%
## Population = AK   6 0.7204 [0.6799; 0.7609] 0.0016 0.0396 13.56 63.1%
## 
## Test for subgroup differences (random effects model):
##                       Q d.f.  p-value
## Between groups   108.86    8 < 0.0001
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
Via_meta <- metaAnalysisRandomModel(Via_effects %>% filter(Sex == "NA"))

p_Via_meta_SE <- Via_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) + 
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Via_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Via_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_Via_meta_CI <- Via_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Age at death (days)", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Via_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Via_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")

p_Via_meta <- ggarrange(p_Via_meta_SE, p_Via_meta_CI)
p_Via_meta

Developmental time

Egg-to-adult Developmental time

DT_A_effects <- DT_A_fitted %>%
  dplyr::select(-Value) %>%
  mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
         Lab = as.factor(Lab),
         V = SE^2,
         Study = paste(Population, Lab, sep = "_")) %>%
  relocate(Population, Sex, Lab, Study) %>%
  arrange(Population) %>% 
  dplyr::rename(Y = Estimate)

Females

  # run the analysis with metagen package
  DT_A_F_metagen <- metagen(data = DT_A_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  DT_A_F_metagen <- update.meta(DT_A_F_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(DT_A_F_metagen), file = "DevelopmentTime/DT_A_F_metagen_sum.txt")
  DT_A_F_metagen

## Number of studies combined: k = 48
## 
##                           SMD               95%-CI     z p-value
## Random effects model 235.0250 [229.0286; 241.0213] 76.82       0
## 
## Quantifying heterogeneity:
##  tau^2 = 438.3997 [318.7134; 790.2616]; tau = 20.9380 [17.8525; 28.1116]
##  I^2 = 99.0% [98.9%; 99.1%]; H = 10.26 [9.75; 10.79]
## 
## Test of heterogeneity:
##        Q d.f. p-value
##  4944.28   47       0
## 
## Results for subgroups (random effects model):
##                   k      SMD               95%-CI     tau^2     tau       Q
## Population = YE   5 226.6241 [218.4126; 234.8356]   78.1653  8.8411   68.56
## Population = RE   6 249.7435 [227.5006; 271.9865]  760.1497 27.5708  611.67
## Population = GI   5 231.9296 [224.7327; 239.1265]   54.7206  7.3973   40.64
## Population = MU   6 239.9889 [210.3381; 269.6398] 1364.1978 36.9350 1638.28
## Population = MA   5 229.4643 [222.6121; 236.3164]   51.8470  7.2005   47.79
## Population = UM   5 230.3775 [223.4138; 237.3412]   52.8420  7.2693   44.77
## Population = KA   5 225.6804 [221.1081; 230.2527]   19.2827  4.3912   20.38
## Population = VA   5 233.7683 [226.7707; 240.7659]   54.4184  7.3769   49.83
## Population = AK   6 240.0130 [208.5734; 271.4527] 1534.8028 39.1766 1883.98
##                   I^2
## Population = YE 94.2%
## Population = RE 99.2%
## Population = GI 90.2%
## Population = MU 99.7%
## Population = MA 91.6%
## Population = UM 91.1%
## Population = KA 80.4%
## Population = VA 92.0%
## Population = AK 99.7%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   9.17    8  0.3282
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  DT_A_F_meta <- metaAnalysisRandomModel(DT_A_effects %>% filter(Sex == "F"))

  p_DT_A_F_meta_SE <- DT_A_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", DT_A_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", DT_A_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_DT_A_F_meta_CI <- DT_A_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", DT_A_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", DT_A_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_DT_A_F_meta <- ggarrange(p_DT_A_F_meta_SE, p_DT_A_F_meta_CI)
  p_DT_A_F_meta

Males

  # run the analysis with metagen package
  DT_A_M_metagen <- metagen(data = DT_A_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  DT_A_M_metagen <- update.meta(DT_A_M_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(DT_A_M_metagen), file = "DevelopmentTime/DT_A_M_metagen_sum.txt")
  DT_A_M_metagen

## Number of studies combined: k = 48
## 
##                           SMD               95%-CI     z p-value
## Random effects model 239.9937 [234.8615; 245.1259] 91.65       0
## 
## Quantifying heterogeneity:
##  tau^2 = 317.5457 [280.5629; 698.4731]; tau = 17.8198 [16.7500; 26.4286]
##  I^2 = 98.6% [98.4%; 98.7%]; H = 8.37 [7.90; 8.87]
## 
## Test of heterogeneity:
##        Q d.f. p-value
##  3294.35   47       0
## 
## Results for subgroups (random effects model):
##                   k      SMD               95%-CI     tau^2     tau       Q
## Population = YE   5 231.3566 [222.3748; 240.3383]   94.7808  9.7355   77.48
## Population = RE   6 252.9210 [232.7427; 273.0992]  621.7322 24.9346  457.91
## Population = GI   5 239.3670 [231.3226; 247.4113]   70.4285  8.3922   48.53
## Population = MU   6 245.5219 [219.6757; 271.3680] 1033.4373 32.1471 1048.60
## Population = MA   5 233.9986 [225.7460; 242.2511]   78.5414  8.8624   66.63
## Population = UM   5 235.2384 [227.6374; 242.8395]   64.1779  8.0111   50.69
## Population = KA   5 229.4524 [224.0212; 234.8836]   29.4632  5.4280   27.59
## Population = VA   5 239.1019 [231.0362; 247.1675]   74.5193  8.6325   63.14
## Population = AK   6 246.2495 [218.6256; 273.8735] 1182.0341 34.3807 1213.95
##                   I^2
## Population = YE 94.8%
## Population = RE 98.9%
## Population = GI 91.8%
## Population = MU 99.5%
## Population = MA 94.0%
## Population = UM 92.1%
## Population = KA 85.5%
## Population = VA 93.7%
## Population = AK 99.6%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   11.08    8  0.1972
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  DT_A_M_meta <- metaAnalysisRandomModel(DT_A_effects %>% filter(Sex == "M"))

  p_DT_A_M_meta_SE <- DT_A_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", DT_A_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", DT_A_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_DT_A_M_meta_CI <- DT_A_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", DT_A_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", DT_A_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_DT_A_M_meta <- ggarrange(p_DT_A_M_meta_SE, p_DT_A_M_meta_CI)
  p_DT_A_M_meta

Dry weight

  DW_effects <- DW_fitted %>%
    dplyr::select(-Value) %>%
    mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
           Lab = as.factor(Lab),
           V = SE^2,
           Study = paste(Population, Lab, sep = "_")) %>%
    relocate(Population, Sex, Lab, Study) %>%
    arrange(Population) %>% 
    dplyr::rename(Y = Estimate)

Females

  # run the analysis with metagen package
  DW_F_metagen <- metagen(data = DW_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  DW_F_metagen <- update.meta(DW_F_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(DW_F_metagen), file = "DryWeight/DW_F_metagen_sum.txt")
  DW_F_metagen

## Number of studies combined: k = 27
## 
##                         SMD           95%-CI      z p-value
## Random effects model 0.4646 [0.4562; 0.4731] 107.39       0
## 
## Quantifying heterogeneity:
##  tau^2 = 0.0004 [0.0002; 0.0009]; tau = 0.0196 [0.0135; 0.0294]
##  I^2 = 76.9% [66.8%; 84.0%]; H = 2.08 [1.73; 2.50]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  112.72   26 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k    SMD           95%-CI   tau^2    tau     Q   I^2
## Population = YE   3 0.4502 [0.4219; 0.4785]  0.0005 0.0226 11.01 81.8%
## Population = RE   3 0.4824 [0.4537; 0.5111]  0.0005 0.0225  9.59 79.2%
## Population = GI   3 0.4545 [0.4321; 0.4769]  0.0003 0.0160  5.77 65.3%
## Population = MU   3 0.4599 [0.4364; 0.4834]  0.0003 0.0179  7.78 74.3%
## Population = MA   3 0.4702 [0.4387; 0.5017]  0.0007 0.0257 13.79 85.5%
## Population = UM   3 0.4705 [0.4423; 0.4988]  0.0005 0.0223  9.98 80.0%
## Population = KA   3 0.4714 [0.4481; 0.4948]  0.0003 0.0177  7.54 73.5%
## Population = VA   3 0.4794 [0.4653; 0.4935] <0.0001 0.0066  2.78 27.9%
## Population = AK   3 0.4431 [0.4195; 0.4666]  0.0003 0.0179  7.76 74.2%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   11.21    8  0.1903
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  DW_F_meta <- metaAnalysisRandomModel(DW_effects %>% filter(Sex == "F"))

  p_DW_F_meta_SE <- DW_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", DW_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", DW_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_DW_F_meta_CI <- DW_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", DW_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", DW_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_DW_F_meta <- ggarrange(p_DW_F_meta_SE, p_DW_F_meta_CI)
  p_DW_F_meta

Males

  # run the analysis with metagen package
  DW_M_metagen <- metagen(data = DW_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  DW_M_metagen <- update.meta(DW_M_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(DW_M_metagen), file = "DryWeight/DW_M_metagen_sum.txt")
  DW_M_metagen

## Number of studies combined: k = 27
## 
##                         SMD           95%-CI      z p-value
## Random effects model 0.2661 [0.2629; 0.2693] 164.95       0
## 
## Quantifying heterogeneity:
##  tau^2 < 0.0001 [0.0000; 0.0001]; tau = 0.0060 [0.0000; 0.0096]
##  I^2 = 53.6% [28.1%; 70.0%]; H = 1.47 [1.18; 1.83]
## 
## Test of heterogeneity:
##      Q d.f. p-value
##  56.01   26  0.0006
## 
## Results for subgroups (random effects model):
##                   k    SMD           95%-CI   tau^2    tau    Q   I^2
## Population = YE   3 0.2547 [0.2454; 0.2640] <0.0001 0.0059 4.15 51.8%
## Population = RE   3 0.2717 [0.2648; 0.2785]       0      0 1.79  0.0%
## Population = GI   3 0.2595 [0.2521; 0.2669] <0.0001 0.0023 2.27 12.1%
## Population = MU   3 0.2690 [0.2629; 0.2752]       0      0 1.71  0.0%
## Population = MA   3 0.2702 [0.2640; 0.2764]       0      0 0.39  0.0%
## Population = UM   3 0.2713 [0.2647; 0.2778]       0      0 1.98  0.0%
## Population = KA   3 0.2731 [0.2638; 0.2824] <0.0001 0.0060 4.24 52.9%
## Population = VA   3 0.2684 [0.2623; 0.2746]       0      0 0.52  0.0%
## Population = AK   3 0.2595 [0.2533; 0.2656]       0      0 0.88  0.0%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   23.04    8  0.0033
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  DW_M_meta <- metaAnalysisRandomModel(DW_effects %>% filter(Sex == "M"))

  p_DW_M_meta_SE <- DW_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", DW_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", DW_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_DW_M_meta_CI <- DW_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", DW_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", DW_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_DW_M_meta <- ggarrange(p_DW_M_meta_SE, p_DW_M_meta_CI)
  p_DW_M_meta

Thorax length

  TL_effects <- TL_fitted %>%
    dplyr::select(-Value) %>%
    mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
           Lab = as.factor(Lab),
           V = SE^2,
           Study = paste(Population, Lab, sep = "_")) %>%
    relocate(Population, Sex, Lab, Study) %>%
    arrange(Population) %>% 
    dplyr::rename(Y = Estimate)

Females

  # run the analysis with metagen package
  TL_F_metagen <- metagen(data = TL_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  TL_F_metagen <- update.meta(TL_F_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(TL_F_metagen), file = "ThoraxLength/TL_F_metagen_sum.txt")
  TL_F_metagen

## Number of studies combined: k = 33
## 
##                           SMD               95%-CI      z p-value
## Random effects model 960.0832 [946.7718; 973.3945] 141.36       0
## 
## Quantifying heterogeneity:
##  tau^2 = 1324.6357 [1037.1888; 3338.0798]; tau = 36.3955 [32.2054; 57.7761]
##  I^2 = 94.0% [92.5%; 95.2%]; H = 4.09 [3.66; 4.57]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  534.55   32 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k      SMD                95%-CI     tau^2     tau     Q
## Population = YE   4 974.1657 [936.4268; 1011.9046] 1286.6997 35.8706 32.40
## Population = RE   3 969.1048 [896.5261; 1041.6834] 3935.7134 62.7353 65.54
## Population = GI   4 968.7977 [918.0887; 1019.5067] 2454.4199 49.5421 47.17
## Population = MU   4 947.3885 [892.5535; 1002.2235] 2861.4591 53.4926 70.74
## Population = MA   4 961.8267 [913.9376; 1009.7159] 2186.5918 46.7610 57.59
## Population = UM   3 955.5090 [893.8572; 1017.1608] 2794.5276 52.8633 48.68
## Population = KA   4 951.2364 [895.4286; 1007.0443] 3031.8513 55.0622 70.45
## Population = VA   3 959.1809 [904.6044; 1013.7574] 2170.5042 46.5887 49.62
## Population = AK   4 943.8390 [897.2902;  990.3877] 2047.5637 45.2500 56.53
##                   I^2
## Population = YE 90.7%
## Population = RE 96.9%
## Population = GI 93.6%
## Population = MU 95.8%
## Population = MA 94.8%
## Population = UM 95.9%
## Population = KA 95.7%
## Population = VA 96.0%
## Population = AK 94.7%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   1.51    8  0.9926
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  TL_F_meta <- metaAnalysisRandomModel(TL_effects %>% filter(Sex == "F"))

  p_TL_F_meta_SE <- TL_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", TL_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", TL_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_TL_F_meta_CI <- TL_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", TL_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", TL_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_TL_F_meta <- ggarrange(p_TL_F_meta_SE, p_TL_F_meta_CI)
  p_TL_F_meta

Males

# run the analysis with metagen package
TL_M_metagen <- metagen(data = TL_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
TL_M_metagen <- update.meta(TL_M_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(TL_M_metagen), file = "ThoraxLength/TL_M_metagen_sum.txt")
TL_M_metagen

## Number of studies combined: k = 24
## 
##                           SMD               95%-CI      z p-value
## Random effects model 827.3092 [816.9626; 837.6559] 156.72       0
## 
## Quantifying heterogeneity:
##  tau^2 = 430.6821 [188.4353; 1136.5359]; tau = 20.7529 [13.7272; 33.7125]
##  I^2 = 71.2% [56.6%; 80.9%]; H = 1.86 [1.52; 2.29]
## 
## Test of heterogeneity:
##      Q d.f.  p-value
##  79.91   23 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k      SMD               95%-CI     tau^2     tau     Q   I^2
## Population = YE   3 840.4294 [824.7127; 856.1460]         0       0  1.76  0.0%
## Population = RE   2 833.0005 [777.0041; 888.9969] 1428.7483 37.7988  7.80 87.2%
## Population = GI   3 832.2111 [814.6964; 849.7258]         0       0  0.47  0.0%
## Population = MU   3 812.3640 [755.2846; 869.4435] 2189.1779 46.7887 17.69 88.7%
## Population = MA   3 830.8742 [782.2636; 879.4847] 1581.2339 39.7647 15.37 87.0%
## Population = UM   2 834.4763 [817.0744; 851.8782]         0       0  0.68  0.0%
## Population = KA   3 815.2189 [799.3232; 831.1145]         0       0  0.29  0.0%
## Population = VA   2 829.5396 [769.1055; 889.9737] 1719.1689 41.4629 10.04 90.0%
## Population = AK   3 814.4255 [777.2659; 851.5850]  816.1319 28.5680  8.81 77.3%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   6.36    8  0.6069
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
TL_M_meta <- metaAnalysisRandomModel(TL_effects %>% filter(Sex == "M"))

p_TL_M_meta_SE <- TL_M_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", TL_M_meta$Qtest_het$Q), 
           hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", TL_M_meta$Qtest_het$p), 
           hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_TL_M_meta_CI <- TL_M_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", TL_M_meta$Qtest_het$Q), 
           hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", TL_M_meta$Qtest_het$p), 
           hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
p_TL_M_meta <- ggarrange(p_TL_M_meta_SE, p_TL_M_meta_CI)
p_TL_M_meta

Wing area

Left

WA_L_effects <- WA_L_fitted %>%
  dplyr::select(-Value) %>%
  mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
         Lab = as.factor(Lab),
         V = SE^2,
         Study = paste(Population, Lab, sep = "_")) %>%
  relocate(Population, Sex, Lab, Study) %>%
  arrange(Population) %>% 
  dplyr::rename(Y = Estimate)

Females

# run the analysis with metagen package
WA_L_F_metagen <- metagen(data = WA_L_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
WA_L_F_metagen <- update.meta(WA_L_F_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(WA_L_F_metagen), file = "WingArea/WA_L_F_metagen_sum.txt")
WA_L_F_metagen

## Number of studies combined: k = 36
## 
##                            SMD                 95%-CI      z p-value
## Random effects model 2800.8679 [2777.4594; 2824.2765] 234.51       0
## 
## Quantifying heterogeneity:
##  tau^2 = 3690.5541 [3973.3996; 13050.5557]; tau = 60.7499 [63.0349; 114.2390]
##  I^2 = 88.2% [84.7%; 90.9%]; H = 2.91 [2.56; 3.32]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  296.72   35 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k       SMD                 95%-CI     tau^2     tau     Q
## Population = YE   4 2791.3037 [2742.4298; 2840.1776] 1392.4646 37.3157  9.15
## Population = RE   4 2888.4516 [2822.3545; 2954.5487] 3148.0819 56.1078 15.91
## Population = GI   4 2769.3180 [2688.7777; 2849.8582] 5329.8130 73.0056 23.35
## Population = MU   4 2779.4302 [2696.0266; 2862.8337] 5691.8809 75.4446 29.16
## Population = MA   4 2831.2333 [2760.7056; 2901.7610] 3722.2356 61.0101 18.23
## Population = UM   4 2847.0409 [2778.5889; 2915.4929] 3459.4232 58.8169 17.79
## Population = KA   4 2779.6403 [2711.7830; 2847.4976] 3318.5528 57.6069 18.05
## Population = VA   4 2766.8715 [2682.9543; 2850.7888] 5774.5174 75.9902 29.24
## Population = AK   4 2753.0461 [2689.0424; 2817.0499] 2851.7529 53.4018 16.00
##                   I^2
## Population = YE 67.2%
## Population = RE 81.1%
## Population = GI 87.1%
## Population = MU 89.7%
## Population = MA 83.5%
## Population = UM 83.1%
## Population = KA 83.4%
## Population = VA 89.7%
## Population = AK 81.2%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   13.33    8  0.1010
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
WA_L_F_meta <- metaAnalysisRandomModel(WA_L_effects %>% filter(Sex == "F"))

p_WA_L_F_meta_SE <- WA_L_F_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", WA_L_F_meta$Qtest_het$Q), 
           hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", WA_L_F_meta$Qtest_het$p), 
           hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_WA_L_F_meta_CI <- WA_L_F_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) + labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", WA_L_F_meta$Qtest_het$Q), 
           hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", WA_L_F_meta$Qtest_het$p), 
           hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
p_WA_L_F_meta <- ggarrange(p_WA_L_F_meta_SE, p_WA_L_F_meta_CI)
p_WA_L_F_meta

Males

# run the analysis with metagen package
WA_L_M_metagen <- metagen(data = WA_L_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
WA_L_M_metagen <- update.meta(WA_L_M_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(WA_L_M_metagen), file = "WingArea/WA_L_M_metagen_sum.txt")
WA_L_M_metagen

## Number of studies combined: k = 36
## 
##                            SMD                 95%-CI      z p-value
## Random effects model 2453.1892 [2437.3661; 2469.0123] 303.87       0
## 
## Quantifying heterogeneity:
##  tau^2 = 1554.6292 [1332.3102; 4689.5239]; tau = 39.4288 [36.5008; 68.4801]
##  I^2 = 82.9% [77.2%; 87.2%]; H = 2.42 [2.09; 2.80]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  205.12   35 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k       SMD                 95%-CI     tau^2     tau     Q
## Population = YE   4 2436.6686 [2413.4651; 2459.8721]  154.6226 12.4347  4.14
## Population = RE   4 2510.8246 [2466.2599; 2555.3892] 1304.8650 36.1229 11.78
## Population = GI   4 2457.5696 [2425.6232; 2489.5159]  456.7954 21.3728  5.73
## Population = MU   4 2434.6340 [2383.6494; 2485.6185] 1895.7661 43.5404 17.68
## Population = MA   4 2485.8104 [2452.7715; 2518.8492]  540.4270 23.2471  6.64
## Population = UM   4 2490.2166 [2449.2856; 2531.1477] 1038.1156 32.2198 10.37
## Population = KA   4 2435.2452 [2392.8096; 2477.6808] 1163.5061 34.1102 12.04
## Population = VA   4 2452.5836 [2413.4794; 2491.6878]  925.1614 30.4165 10.07
## Population = AK   4 2417.1283 [2379.3148; 2454.9417]  820.0968 28.6373  9.28
##                   I^2
## Population = YE 27.5%
## Population = RE 74.5%
## Population = GI 47.7%
## Population = MU 83.0%
## Population = MA 54.8%
## Population = UM 71.1%
## Population = KA 75.1%
## Population = VA 70.2%
## Population = AK 67.7%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   19.95    8  0.0105
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
WA_L_M_meta <- metaAnalysisRandomModel(WA_L_effects %>% filter(Sex == "M"))

p_WA_L_M_meta_SE <- WA_L_M_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", WA_L_M_meta$Qtest_het$Q), 
           hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", WA_L_M_meta$Qtest_het$p), 
           hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_WA_L_M_meta_CI <- WA_L_M_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", WA_L_M_meta$Qtest_het$Q), 
           hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", WA_L_M_meta$Qtest_het$p), 
           hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
p_WA_L_M_meta <- ggarrange(p_WA_L_M_meta_SE, p_WA_L_M_meta_CI)
p_WA_L_M_meta

Right

  WA_R_effects <- WA_R_fitted %>%
    dplyr::select(-Value) %>%
    mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
           Lab = as.factor(Lab),
           V = SE^2,
           Study = paste(Population, Lab, sep = "_")) %>%
    relocate(Population, Sex, Lab, Study) %>%
    arrange(Population) %>% 
    dplyr::rename(Y = Estimate)

Females

  # run the analysis with metagen package
  WA_R_F_metagen <- metagen(data = WA_R_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  WA_R_F_metagen <- update.meta(WA_R_F_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(WA_R_F_metagen), file = "WingArea/WA_R_F_metagen_sum.txt")
  WA_R_F_metagen

## Number of studies combined: k = 36
## 
##                            SMD                 95%-CI      z p-value
## Random effects model 2804.3284 [2780.8876; 2827.7692] 234.48       0
## 
## Quantifying heterogeneity:
##  tau^2 = 3712.2491 [3978.0522; 13037.7622]; tau = 60.9282 [63.0718; 114.1830]
##  I^2 = 88.4% [84.9%; 91.0%]; H = 2.93 [2.57; 3.34]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  300.76   35 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k       SMD                 95%-CI     tau^2     tau     Q
## Population = YE   4 2794.0556 [2745.8689; 2842.2422] 1344.1037 36.6620  9.07
## Population = RE   4 2891.1551 [2825.2956; 2957.0147] 3127.1890 55.9213 16.10
## Population = GI   4 2775.6368 [2700.5171; 2850.7564] 4493.1225 67.0308 20.46
## Population = MU   4 2784.8949 [2702.7552; 2867.0346] 5494.3336 74.1238 28.89
## Population = MA   4 2836.0644 [2765.7257; 2906.4032] 3705.3807 60.8718 18.47
## Population = UM   4 2852.1207 [2783.1491; 2921.0922] 3536.3308 59.4671 18.47
## Population = KA   4 2783.7269 [2716.7682; 2850.6855] 3214.8042 56.6992 17.95
## Population = VA   4 2770.7951 [2688.3713; 2853.2190] 5537.9221 74.4172 28.80
## Population = AK   4 2753.1090 [2688.0545; 2818.1635] 2983.7256 54.6235 16.95
##                   I^2
## Population = YE 66.9%
## Population = RE 81.4%
## Population = GI 85.3%
## Population = MU 89.6%
## Population = MA 83.8%
## Population = UM 83.8%
## Population = KA 83.3%
## Population = VA 89.6%
## Population = AK 82.3%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   13.62    8  0.0922
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  WA_R_F_meta <- metaAnalysisRandomModel(WA_R_effects %>% filter(Sex == "F"))

  p_WA_R_F_meta_SE <- WA_R_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", WA_R_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", WA_R_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_WA_R_F_meta_CI <- WA_R_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", WA_R_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", WA_R_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_WA_R_F_meta <- ggarrange(p_WA_R_F_meta_SE, p_WA_R_F_meta_CI)
  p_WA_R_F_meta

Males

  # run the analysis with metagen package
  WA_R_M_metagen <- metagen(data = WA_R_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  WA_R_M_metagen <- update.meta(WA_R_M_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(WA_R_M_metagen), file = "WingArea/WA_R_M_metagen_sum.txt")
  WA_R_M_metagen

## Number of studies combined: k = 36
## 
##                            SMD                 95%-CI      z p-value
## Random effects model 2457.1161 [2441.2244; 2473.0078] 303.04       0
## 
## Quantifying heterogeneity:
##  tau^2 = 1571.1467 [1357.0746; 4762.2285]; tau = 39.6377 [36.8385; 69.0089]
##  I^2 = 83.0% [77.2%; 87.2%]; H = 2.42 [2.10; 2.80]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  205.34   35 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k       SMD                 95%-CI     tau^2     tau     Q
## Population = YE   4 2438.9400 [2415.8051; 2462.0748]  151.3581 12.3028  4.11
## Population = RE   4 2516.9424 [2473.3365; 2560.5484] 1226.9858 35.0284 11.21
## Population = GI   4 2461.1887 [2433.3782; 2488.9991]  267.7837 16.3641  4.59
## Population = MU   4 2442.1341 [2393.7544; 2490.5138] 1652.2214 40.6475 15.77
## Population = MA   4 2488.7826 [2456.3000; 2521.2652]  510.3641 22.5912  6.42
## Population = UM   4 2494.9835 [2453.6122; 2536.3548] 1068.4866 32.6877 10.55
## Population = KA   4 2439.1817 [2395.5320; 2482.8313] 1255.7049 35.4359 12.75
## Population = VA   4 2455.4881 [2416.3765; 2494.5997]  925.0574 30.4148 10.06
## Population = AK   4 2418.3341 [2378.6557; 2458.0124]  938.6895 30.6380 10.19
##                   I^2
## Population = YE 27.0%
## Population = RE 73.2%
## Population = GI 34.6%
## Population = MU 81.0%
## Population = MA 53.2%
## Population = UM 71.6%
## Population = KA 76.5%
## Population = VA 70.2%
## Population = AK 70.5%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   21.15    8  0.0067
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  WA_R_M_meta <- metaAnalysisRandomModel(WA_R_effects %>% filter(Sex == "M"))

  p_WA_R_M_meta_SE <- WA_R_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", WA_R_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", WA_R_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_WA_R_M_meta_CI <- WA_R_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", WA_R_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", WA_R_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_WA_R_M_meta <- ggarrange(p_WA_R_M_meta_SE, p_WA_R_M_meta_CI)
  p_WA_R_M_meta

Fecundity

  Fec_effects <- Fec_fitted %>%
    dplyr::select(-Value) %>%
    mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
           Lab = as.factor(Lab),
           V = SE^2,
           Study = paste(Population, Lab, sep = "_")) %>%
    relocate(Population, Sex, Lab, Study) %>%
    arrange(Population) %>% 
    dplyr::rename(Y = Estimate)

  # run the analysis with metagen package
  Fec_metagen <- metagen(data = Fec_effects %>% filter(Sex == "NA"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  Fec_metagen <- update.meta(Fec_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(Fec_metagen), file = "WingArea/Fec_metagen_sum.txt")
  Fec_metagen

## Number of studies combined: k = 18
## 
##                          SMD             95%-CI     z  p-value
## Random effects model 70.1795 [61.1053; 79.2538] 15.16 < 0.0001
## 
## Quantifying heterogeneity:
##  tau^2 = 316.4696 [169.2196; 883.0771]; tau = 17.7896 [13.0084; 29.7166]
##  I^2 = 83.5% [75.2%; 89.1%]; H = 2.46 [2.01; 3.02]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  103.20   17 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k     SMD              95%-CI     tau^2     tau     Q   I^2
## Population = YE   2 56.6295 [46.1185;  67.1405]         0       0  0.96  0.0%
## Population = RE   2 60.6816 [48.7946;  72.5686]         0       0  0.63  0.0%
## Population = GI   2 61.1849 [38.0683;  84.3016]  198.8189 14.1003  3.42 70.8%
## Population = MU   2 74.7237 [38.5098; 110.9375]  615.7878 24.8151 10.08 90.1%
## Population = MA   2 76.9132 [34.6604; 119.1660]  860.4953 29.3342 13.33 92.5%
## Population = UM   2 74.3382 [19.6394; 129.0370] 1479.4131 38.4631 19.73 94.9%
## Population = KA   2 80.0387 [22.0330; 138.0445] 1688.4587 41.0909 27.58 96.4%
## Population = VA   2 69.5147 [37.7343; 101.2951]  463.3340 21.5252  8.33 88.0%
## Population = AK   2 78.5449 [43.2876; 113.8023]  582.8642 24.1426  9.98 90.0%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   3.61    8  0.8902
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  Fec_meta <- metaAnalysisRandomModel(Fec_effects %>% filter(Sex == "NA"))

  p_Fec_meta_SE <- Fec_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Fec_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Fec_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_Fec_meta_CI <- Fec_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Fec_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Fec_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_Fec_meta <- ggarrange(p_Fec_meta_SE, p_Fec_meta_CI)
  p_Fec_meta

Lifespan

  LS_effects <- LS_fitted %>%
    dplyr::select(-Value) %>%
    mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
           Lab = as.factor(Lab),
           V = SE^2,
           Study = paste(Population, Lab, sep = "_")) %>%
    relocate(Population, Sex, Lab, Study) %>%
    arrange(Population) %>% 
    dplyr::rename(Y = Estimate)

Females

  # run the analysis with metagen package
  LS_F_metagen <- metagen(data = LS_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  LS_F_metagen <- update.meta(LS_F_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(LS_F_metagen), file = "Lifespan/LS_F_metagen_sum.txt")
  LS_F_metagen

## Number of studies combined: k = 27
## 
##                          SMD             95%-CI     z p-value
## Random effects model 49.5126 [47.1345; 51.8907] 40.81       0
## 
## Quantifying heterogeneity:
##  tau^2 = 35.7209 [19.9491; 71.9771]; tau = 5.9767 [4.4664; 8.4839]
##  I^2 = 95.0% [93.7%; 96.1%]; H = 4.48 [3.99; 5.04]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  522.79   26 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k     SMD             95%-CI   tau^2    tau     Q   I^2
## Population = YE   3 48.2235 [41.6119; 54.8351] 30.2032 5.4957 23.22 91.4%
## Population = RE   3 45.4036 [43.8697; 46.9375]       0      0  0.62  0.0%
## Population = GI   3 46.5748 [40.9487; 52.2010] 20.6103 4.5399 14.55 86.3%
## Population = MU   3 51.7848 [42.3130; 61.2566] 65.9619 8.1217 48.17 95.8%
## Population = MA   3 47.2862 [42.3160; 52.2564] 15.5781 3.9469 12.91 84.5%
## Population = UM   3 46.6874 [38.0726; 55.3023] 53.8217 7.3363 38.27 94.8%
## Population = KA   3 52.8596 [46.5610; 59.1581] 27.0879 5.2046 21.07 90.5%
## Population = VA   3 52.1671 [43.5001; 60.8341] 54.6033 7.3894 40.25 95.0%
## Population = AK   3 54.4827 [43.1011; 65.8643] 97.0692 9.8524 71.88 97.2%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   10.96    8  0.2040
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  LS_F_meta <- metaAnalysisRandomModel(LS_effects %>% filter(Sex == "F"))

  p_LS_F_meta_SE <- LS_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", LS_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", LS_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_LS_F_meta_CI <- LS_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", LS_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", LS_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_LS_F_meta <- ggarrange(p_LS_F_meta_SE, p_LS_F_meta_CI)
  p_LS_F_meta

Males

# run the analysis with metagen package
LS_M_metagen <- metagen(data = LS_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
LS_M_metagen <- update.meta(LS_M_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(LS_M_metagen), file = "Lifespan/LS_M_metagen_sum.txt")
LS_M_metagen

## Number of studies combined: k = 27
## 
##                          SMD             95%-CI     z p-value
## Random effects model 52.0083 [50.7576; 53.2590] 81.50       0
## 
## Quantifying heterogeneity:
##  tau^2 = 8.5327 [6.3041; 24.3995]; tau = 2.9211 [2.5108; 4.9396]
##  I^2 = 84.3% [78.2%; 88.7%]; H = 2.52 [2.14; 2.97]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  165.46   26 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k     SMD             95%-CI   tau^2    tau     Q   I^2
## Population = YE   3 49.5429 [44.4855; 54.6003] 17.4778 4.1806 18.00 88.9%
## Population = RE   3 49.9381 [48.5232; 51.3530]       0      0  1.26  0.0%
## Population = GI   3 50.3927 [48.7202; 52.0652]  0.3588 0.5990  2.28 12.4%
## Population = MU   3 54.8573 [50.9592; 58.7553]  9.4434 3.0730 10.61 81.1%
## Population = MA   3 50.6985 [47.1443; 54.2528]  7.4838 2.7357  8.86 77.4%
## Population = UM   3 50.9024 [44.4279; 57.3768] 30.1193 5.4881 28.55 93.0%
## Population = KA   3 55.6251 [54.0750; 57.1753]  0.2287 0.4782  2.21  9.5%
## Population = VA   3 53.4606 [48.6456; 58.2756] 15.6045 3.9503 16.11 87.6%
## Population = AK   3 53.9160 [49.0280; 58.8039] 16.2146 4.0267 17.17 88.4%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f.  p-value
## Between groups   37.25    8 < 0.0001
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
LS_M_meta <- metaAnalysisRandomModel(LS_effects %>% filter(Sex == "M"))

p_LS_M_meta_SE <- LS_M_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", LS_M_meta$Qtest_het$Q), hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", LS_M_meta$Qtest_het$p), hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_LS_M_meta_CI <- LS_M_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", LS_M_meta$Qtest_het$Q), hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", LS_M_meta$Qtest_het$p), hjust=-0.2, vjust=5, parse = T) +
theme(legend.position = "none")
  
p_LS_M_meta <- ggarrange(p_LS_M_meta_SE, p_LS_M_meta_CI)
p_LS_M_meta

Cold-shock mortality

CSM_effects <- CSM_fitted %>%
  dplyr::select(-Value) %>%
  mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
         Lab = as.factor(Lab),
         V = SE^2,
         Study = paste(Population, Lab, sep = "_")) %>%
  relocate(Population, Sex, Lab, Study) %>%
  arrange(Population) %>% 
  dplyr::rename(Y = Estimate)

Females

# run the analysis with metagen package
CSM_F_metagen <- metagen(data = CSM_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
CSM_F_metagen <- update.meta(CSM_F_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(CSM_F_metagen), file = "ColdShock/CSM_F_metagen_sum.txt")
CSM_F_metagen

## Number of studies combined: k = 27
## 
##                         SMD           95%-CI     z p-value
## Random effects model 0.8055 [0.7715; 0.8396] 46.34       0
## 
## Quantifying heterogeneity:
##  tau^2 = 0.0052 [0.0073; 0.0289]; tau = 0.0719 [0.0852; 0.1701]
##  I^2 = 75.1% [63.9%; 82.9%]; H = 2.01 [1.66; 2.42]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  104.61   26 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k    SMD           95%-CI  tau^2    tau     Q   I^2
## Population = YE   3 0.8187 [0.7055; 0.9318] 0.0068 0.0827  8.89 77.5%
## Population = RE   3 0.7911 [0.6714; 0.9109] 0.0077 0.0878  8.63 76.8%
## Population = GI   3 0.7701 [0.6566; 0.8836] 0.0064 0.0803  7.02 71.5%
## Population = MU   3 0.8197 [0.7041; 0.9353] 0.0073 0.0853  9.35 78.6%
## Population = MA   3 0.7891 [0.6670; 0.9113] 0.0084 0.0915 10.49 80.9%
## Population = UM   3 0.7284 [0.6133; 0.8435] 0.0070 0.0834  8.21 75.6%
## Population = KA   3 0.7907 [0.6497; 0.9316] 0.0117 0.1084 13.00 84.6%
## Population = VA   3 0.8010 [0.6536; 0.9484] 0.0133 0.1155 15.39 87.0%
## Population = AK   3 0.8359 [0.7347; 0.9370] 0.0052 0.0720  7.42 73.0%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   2.47    8  0.9633
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
CSM_F_meta <- metaAnalysisRandomModel(CSM_effects %>% filter(Sex == "F"))

p_CSM_F_meta_SE <- CSM_F_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", CSM_F_meta$Qtest_het$Q), hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", CSM_F_meta$Qtest_het$p), hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_CSM_F_meta_CI <- CSM_F_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", CSM_F_meta$Qtest_het$Q), hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", CSM_F_meta$Qtest_het$p), hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
p_CSM_F_meta <- ggarrange(p_CSM_F_meta_SE, p_CSM_F_meta_CI)
p_CSM_F_meta

Males

# run the analysis with metagen package
CSM_M_metagen <- metagen(data = CSM_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
CSM_M_metagen <- update.meta(CSM_M_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(CSM_M_metagen), file = "ColdShock/CSM_M_metagen_sum.txt")
CSM_M_metagen

## Number of studies combined: k = 27
## 
##                         SMD           95%-CI     z  p-value
## Random effects model 0.7004 [0.6629; 0.7379] 36.59 < 0.0001
## 
## Quantifying heterogeneity:
##  tau^2 = 0.0045 [0.0000; 0.0108]; tau = 0.0674 [0.0000; 0.1038]
##  I^2 = 49.2% [20.7%; 67.5%]; H = 1.40 [1.12; 1.75]
## 
## Test of heterogeneity:
##      Q d.f. p-value
##  51.21   26  0.0022
## 
## Results for subgroups (random effects model):
##                   k    SMD           95%-CI  tau^2    tau     Q   I^2
## Population = YE   3 0.8047 [0.6746; 0.9347] 0.0077 0.0876  4.94 59.5%
## Population = RE   3 0.6870 [0.5739; 0.8001] 0.0042 0.0646  3.43 41.7%
## Population = GI   3 0.7004 [0.6171; 0.7837]      0      0  1.56  0.0%
## Population = MU   3 0.7189 [0.5245; 0.9132] 0.0237 0.1540 11.36 82.4%
## Population = MA   3 0.6852 [0.6093; 0.7611]      0      0  1.77  0.0%
## Population = UM   3 0.6281 [0.5343; 0.7220] 0.0019 0.0433  2.69 25.7%
## Population = KA   3 0.7065 [0.6306; 0.7824]      0      0  1.12  0.0%
## Population = VA   3 0.6823 [0.5300; 0.8346] 0.0125 0.1120  6.88 70.9%
## Population = AK   3 0.6796 [0.5797; 0.7794] 0.0031 0.0553  3.26 38.6%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   5.03    8  0.7546
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
CSM_M_meta <- metaAnalysisRandomModel(CSM_effects %>% filter(Sex == "M"))

p_CSM_M_meta_SE <- CSM_M_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", CSM_M_meta$Qtest_het$Q), hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", CSM_M_meta$Qtest_het$p), hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_CSM_M_meta_CI <- CSM_M_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", CSM_M_meta$Qtest_het$Q), hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", CSM_M_meta$Qtest_het$p), hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
p_CSM_M_meta <- ggarrange(p_CSM_M_meta_SE, p_CSM_M_meta_CI)
p_CSM_M_meta

Chill-coma recovery time

Females

Males

Heat-shock mortality

  HSM_effects <- HSM_fitted %>%
    dplyr::select(-Value) %>%
    mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
           Lab = as.factor(Lab),
           V = SE^2,
           Study = paste(Population, Lab, sep = "_")) %>%
    relocate(Population, Sex, Lab, Study) %>%
    arrange(Population) %>% 
    dplyr::rename(Y = Estimate)

Females

  # run the analysis with metagen package
  HSM_F_metagen <- metagen(data = HSM_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  HSM_F_metagen <- update.meta(HSM_F_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(HSM_F_metagen), file = "HeatShock/HSM_F_metagen_sum.txt")
  HSM_F_metagen

## Number of studies combined: k = 18
## 
##                           SMD               95%-CI     z p-value
## Random effects model 361.3025 [351.9100; 370.6949] 75.39       0
## 
## Quantifying heterogeneity:
##  tau^2 = 237.8482 [117.8599; 1004.1068]; tau = 15.4223 [10.8563; 31.6876]
##  I^2 = 77.6% [65.1%; 85.7%]; H = 2.11 [1.69; 2.64]
## 
## Test of heterogeneity:
##      Q d.f.  p-value
##  76.04   17 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k      SMD               95%-CI     tau^2     tau    Q   I^2
## Population = YE   2 342.6754 [294.8420; 390.5088]  943.5247 30.7168 4.06 75.3%
## Population = RE   2 355.3141 [303.4415; 407.1867] 1133.9239 33.6738 4.55 78.0%
## Population = GI   2 334.2288 [321.8623; 346.5953]         0       0 0.02  0.0%
## Population = MU   2 361.8072 [295.4423; 428.1721] 2011.6901 44.8519 7.35 86.4%
## Population = MA   2 391.0997 [344.1103; 438.0891]  888.9120 29.8146 3.66 72.6%
## Population = UM   2 369.8067 [347.2564; 392.3570]  125.9058 11.2208 1.37 27.1%
## Population = KA   2 364.2032 [352.7194; 375.6869]         0       0 0.05  0.0%
## Population = VA   2 360.0392 [344.1934; 375.8850]   41.7888  6.4644 1.12 11.0%
## Population = AK   2 368.9146 [344.4671; 393.3620]  160.9033 12.6848 1.49 32.7%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   19.10    8  0.0144
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  HSM_F_meta <- metaAnalysisRandomModel(HSM_effects %>% filter(Sex == "F"))

  p_HSM_F_meta_SE <- HSM_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", HSM_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", HSM_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_HSM_F_meta_CI <- HSM_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", HSM_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", HSM_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_HSM_F_meta <- ggarrange(p_HSM_F_meta_SE, p_HSM_F_meta_CI)
  p_HSM_F_meta

Males

  # run the analysis with metagen package
  HSM_M_metagen <- metagen(data = HSM_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  HSM_M_metagen <- update.meta(HSM_M_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(HSM_M_metagen), file = "HeatShock/HSM_M_metagen_sum.txt")
  HSM_M_metagen

## Number of studies combined: k = 18
## 
##                           SMD               95%-CI     z p-value
## Random effects model 322.5888 [310.1101; 335.0675] 50.67       0
## 
## Quantifying heterogeneity:
##  tau^2 = 436.0765 [262.0480; 1936.3667]; tau = 20.8824 [16.1879; 44.0042]
##  I^2 = 69.9% [51.2%; 81.4%]; H = 1.82 [1.43; 2.32]
## 
## Test of heterogeneity:
##      Q d.f.  p-value
##  56.42   17 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k      SMD               95%-CI     tau^2     tau    Q   I^2
## Population = YE   2 295.6367 [230.4706; 360.8028] 1858.0030 43.1046 5.80 82.7%
## Population = RE   2 299.3280 [237.0541; 361.6018] 1645.9981 40.5709 5.04 80.2%
## Population = GI   2 278.4127 [258.2987; 298.5267]         0       0 0.08  0.0%
## Population = MU   2 322.9962 [279.5264; 366.4659]  676.5451 26.0105 2.76 63.7%
## Population = MA   2 349.2955 [283.3611; 415.2299] 1908.2794 43.6839 5.89 83.0%
## Population = UM   2 336.4350 [283.8562; 389.0138] 1107.4937 33.2790 3.83 73.9%
## Population = KA   2 328.3924 [310.4407; 346.3441]         0       0 0.22  0.0%
## Population = VA   2 327.7797 [300.3089; 355.2504]  147.8583 12.1597 1.37 27.0%
## Population = AK   2 355.4177 [320.4275; 390.4078]  369.9475 19.2340 1.97 49.2%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   23.13    8  0.0032
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  HSM_M_meta <- metaAnalysisRandomModel(HSM_effects %>% filter(Sex == "M"))

  p_HSM_M_meta_SE <- HSM_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", HSM_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", HSM_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_HSM_M_meta_CI <- HSM_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", HSM_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", HSM_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_HSM_M_meta <- ggarrange(p_HSM_M_meta_SE, p_HSM_M_meta_CI)
  p_HSM_M_meta

Diapause

  # get the effects in the right shape
  Dia_effects <- Dia_fitted %>%
    dplyr::select(-Value) %>%
    mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
           Lab = as.factor(Lab),
           V = SE^2,
           Study = paste(Population, Lab, sep = "_")) %>%
    relocate(Population, Sex, Lab, Study) %>%
    arrange(Population) %>% 
    dplyr::rename(Y = Estimate)

  # run the analysis with metagen package
  Dia_metagen <- metagen(data = Dia_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  Dia_metagen <- update.meta(Dia_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(Dia_metagen), file = "Diapause/Dia_metagen_sum.txt")
  Dia_metagen

## Number of studies combined: k = 27
## 
##                         SMD           95%-CI     z  p-value
## Random effects model 0.7819 [0.7340; 0.8297] 32.02 < 0.0001
## 
## Quantifying heterogeneity:
##  tau^2 = 0.0136 [0.0122; 0.0456]; tau = 0.1164 [0.1103; 0.2136]
##  I^2 = 92.9% [90.8%; 94.5%]; H = 3.75 [3.30; 4.27]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  366.19   26 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k    SMD           95%-CI  tau^2    tau     Q   I^2
## Population = YE   3 0.7851 [0.5785; 0.9918] 0.0310 0.1759 33.88 94.1%
## Population = RE   3 0.7506 [0.4866; 1.0146] 0.0512 0.2263 40.70 95.1%
## Population = GI   3 0.7201 [0.4489; 0.9914] 0.0544 0.2332 44.71 95.5%
## Population = MU   3 0.8095 [0.5951; 1.0240] 0.0335 0.1831 35.87 94.4%
## Population = MA   3 0.7790 [0.5798; 0.9782] 0.0285 0.1689 30.05 93.3%
## Population = UM   3 0.8798 [0.7335; 1.0261] 0.0139 0.1179 13.59 85.3%
## Population = KA   3 0.7577 [0.4953; 1.0201] 0.0514 0.2268 56.53 96.5%
## Population = VA   3 0.8031 [0.6432; 0.9630] 0.0174 0.1319 18.45 89.2%
## Population = AK   3 0.6772 [0.3523; 1.0021] 0.0799 0.2826 78.09 97.4%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   2.33    8  0.9693
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  Dia_meta <- metaAnalysisRandomModel(Dia_effects %>% filter(Sex == "F"))

  p_Dia_meta_SE <- Dia_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Dia_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Dia_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_Dia_meta_CI <- Dia_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Dia_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Dia_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_Dia_meta <- ggarrange(p_Dia_meta_SE, p_Dia_meta_CI)
  p_Dia_meta

Starvation resistance

  SR_effects <- SR_fitted %>%
    dplyr::select(-Value) %>%
    mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
           Lab = as.factor(Lab),
           V = SE^2,
           Study = paste(Population, Lab, sep = "_")) %>%
    relocate(Population, Sex, Lab, Study) %>%
    arrange(Population) %>% 
    dplyr::rename(Y = Estimate)

Females

  # run the analysis with metagen package
  SR_F_metagen <- metagen(data = SR_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  SR_F_metagen <- update.meta(SR_F_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(SR_F_metagen), file = "Starvation/SR_F_metagen_sum.txt")
  SR_F_metagen

## Number of studies combined: k = 27
## 
##                          SMD             95%-CI     z p-value
## Random effects model 92.2150 [88.2592; 96.1708] 45.69       0
## 
## Quantifying heterogeneity:
##  tau^2 = 95.1301 [58.5957; 208.2583]; tau = 9.7535 [7.6548; 14.4312]
##  I^2 = 87.8% [83.5%; 91.0%]; H = 2.86 [2.46; 3.34]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  213.28   26 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k      SMD              95%-CI    tau^2     tau     Q   I^2
## Population = YE   3  79.1118 [69.3841;  88.8395]  59.4425  7.7099 10.58 81.1%
## Population = RE   3  90.3681 [78.2393; 102.4968]  98.6061  9.9301 14.58 86.3%
## Population = GI   3  90.1160 [73.0025; 107.2294] 211.5175 14.5436 27.83 92.8%
## Population = MU   3  98.8544 [88.4272; 109.2816]  71.2725  8.4423 13.10 84.7%
## Population = MA   3  90.7965 [79.3353; 102.2577]  89.1768  9.4433 16.29 87.7%
## Population = UM   3  92.7478 [81.4423; 104.0532]  83.2652  9.1250 12.39 83.9%
## Population = KA   3  94.2444 [81.1793; 107.3095] 118.6246 10.8915 19.05 89.5%
## Population = VA   3 100.3351 [88.4492; 112.2210]  96.0593  9.8010 16.24 87.7%
## Population = AK   3  93.1682 [84.0970; 102.2394]  51.1189  7.1497 10.21 80.4%
## 
## Test for subgroup differences (random effects model):
##                      Q d.f. p-value
## Between groups   10.64    8  0.2229
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  SR_F_meta <- metaAnalysisRandomModel(SR_effects %>% filter(Sex == "F"))

  p_SR_F_meta_SE <- SR_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", SR_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", SR_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_SR_F_meta_CI <- SR_F_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", SR_F_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", SR_F_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_SR_F_meta <- ggarrange(p_SR_F_meta_SE, p_SR_F_meta_CI)
  p_SR_F_meta

Males

  # run the analysis with metagen package
  SR_M_metagen <- metagen(data = SR_effects %>% filter(Sex == "M"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
  SR_M_metagen <- update.meta(SR_M_metagen, subgroup = Population, tau.common = FALSE)
  capture.output(summary(SR_M_metagen), file = "Starvation/SR_M_metagen_sum.txt")
  SR_M_metagen

## Number of studies combined: k = 27
## 
##                          SMD             95%-CI     z p-value
## Random effects model 67.0390 [64.1185; 69.9594] 44.99       0
## 
## Quantifying heterogeneity:
##  tau^2 = 50.9069 [34.4087; 122.9392]; tau = 7.1349 [5.8659; 11.0878]
##  I^2 = 87.1% [82.5%; 90.6%]; H = 2.79 [2.39; 3.26]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  202.04   26 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k     SMD             95%-CI   tau^2    tau     Q   I^2
## Population = YE   3 57.5509 [49.0188; 66.0831] 48.3328 6.9522 14.65 86.4%
## Population = RE   3 65.2671 [57.3446; 73.1895] 39.4277 6.2791 10.88 81.6%
## Population = GI   3 63.5417 [53.6129; 73.4704] 66.6562 8.1643 15.86 87.4%
## Population = MU   3 73.9147 [66.5549; 81.2746] 33.7392 5.8085 10.71 81.3%
## Population = MA   3 66.0302 [56.5611; 75.4993] 61.3987 7.8357 17.99 88.9%
## Population = UM   3 69.1615 [59.9618; 78.3613] 56.4189 7.5113 15.09 86.7%
## Population = KA   3 69.6949 [61.7787; 77.6111] 40.1309 6.3349 12.25 83.7%
## Population = VA   3 70.8085 [60.8687; 80.7482] 68.5145 8.2773 19.86 89.9%
## Population = AK   3 67.6143 [60.8702; 74.3585] 27.1947 5.2149  9.18 78.2%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   9.99    8  0.2660
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

  # run the analysis with EK's code for quick plotting
  SR_M_meta <- metaAnalysisRandomModel(SR_effects %>% filter(Sex == "M"))

  p_SR_M_meta_SE <- SR_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, SEMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with SE") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", SR_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", SR_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  
  p_SR_M_meta_CI <- SR_M_meta$summary_effects_random %>% 
    dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
    distinct() %>%
    ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
    theme_bw() +
    geom_point(size = 4, shape = 15) +
    geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
    scale_y_continuous(name = "Population", breaks = 1:9, 
                       labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
    labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
    theme(panel.grid.major = element_blank(),
          panel.grid.minor = element_blank()) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", SR_M_meta$Qtest_het$Q), 
             hjust=-0.2, vjust=3.5, parse = T) +
    annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", SR_M_meta$Qtest_het$p), 
             hjust=-0.2, vjust=5, parse = T) +
    theme(legend.position = "none")
  
  p_SR_M_meta <- ggarrange(p_SR_M_meta_SE, p_SR_M_meta_CI)
  p_SR_M_meta

Pigmentation

Pgm_TotalPerc_effects <- Pgm_TotalPerc_fitted %>%
  dplyr::select(-Value) %>%
  mutate(Population = factor(Population, levels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")),
         Lab = as.factor(Lab),
         V = SE^2,
         Study = paste(Population, Lab, sep = "_")) %>%
  relocate(Population, Sex, Lab, Study) %>%
  arrange(Population) %>% 
  dplyr::rename(Y = Estimate)

With all three labs

# run the analysis with metagen package
Pgm_TotalPerc_metagen <- metagen(data = Pgm_TotalPerc_effects %>% filter(Sex == "F"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
Pgm_TotalPerc_metagen <- update.meta(Pgm_TotalPerc_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(Pgm_TotalPerc_metagen), file = "Pigmentation/Pgm_TotalPerc_metagen_sum.txt")
Pgm_TotalPerc_metagen

## Number of studies combined: k = 27
## 
##                          SMD             95%-CI     z  p-value
## Random effects model 22.1310 [18.4664; 25.7957] 11.84 < 0.0001
## 
## Quantifying heterogeneity:
##  tau^2 = 92.7608 [59.7248; 185.9682]; tau = 9.6312 [7.7282; 13.6370]
##  I^2 = 98.4% [98.1%; 98.7%]; H = 7.95 [7.33; 8.62]
## 
## Test of heterogeneity:
##        Q d.f. p-value
##  1641.63   26       0
## 
## Results for subgroups (random effects model):
##                   k     SMD             95%-CI    tau^2     tau      Q   I^2
## Population = YE   3 21.9253 [ 8.9080; 34.9427] 130.8658 11.4397 189.65 98.9%
## Population = RE   3 23.1491 [10.3955; 35.9027] 125.1476 11.1869 140.08 98.6%
## Population = GI   3 19.8218 [ 7.0653; 32.5783] 125.0353 11.1819 127.97 98.4%
## Population = MU   3 21.9975 [ 7.8889; 36.1060] 153.9193 12.4064 211.85 99.1%
## Population = MA   3 26.0765 [11.3607; 40.7923] 167.5581 12.9444 226.86 99.1%
## Population = UM   3 24.2945 [12.6483; 35.9406] 104.2371 10.2097 129.73 98.5%
## Population = KA   3 20.6854 [ 8.4856; 32.8852] 114.7403 10.7117 162.20 98.8%
## Population = VA   3 20.2816 [ 7.0677; 33.4956] 134.7945 11.6101 180.83 98.9%
## Population = AK   3 21.0641 [ 7.6242; 34.5041] 139.6157 11.8159 204.34 99.0%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   0.71    8  0.9995
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
Pgm_TotalPerc_meta <- metaAnalysisRandomModel(Pgm_TotalPerc_effects %>% filter(Sex == "F"))

p_Pgm_TotalPerc_meta_SE <- Pgm_TotalPerc_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Pgm_TotalPerc_meta$Qtest_het$Q),
             hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Pgm_TotalPerc_meta$Qtest_het$p),
             hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_Pgm_TotalPerc_meta_CI <- Pgm_TotalPerc_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Pgm_TotalPerc_meta$Qtest_het$Q),
             hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Pgm_TotalPerc_meta$Qtest_het$p),
             hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
p_Pgm_TotalPerc_meta <- ggarrange(p_Pgm_TotalPerc_meta_SE, p_Pgm_TotalPerc_meta_CI)
p_Pgm_TotalPerc_meta

Without PS data

# run the analysis with metagen package
Pgm_TotalPercwoSchmidt_metagen <- metagen(data = Pgm_TotalPerc_effects %>% filter(Sex == "F" & Lab != "Schmidt"), TE = Y, seTE = SE, studlab = Study, sm = "SMD", fixed = FALSE, random = TRUE, method.tau = "DL")
Pgm_TotalPercwoSchmidt_metagen <- update.meta(Pgm_TotalPercwoSchmidt_metagen, subgroup = Population, tau.common = FALSE)
capture.output(summary(Pgm_TotalPercwoSchmidt_metagen), file = "Pigmentation/Pgm_TotalPercwoSchmidt_metagen_sum.txt")
Pgm_TotalPercwoSchmidt_metagen

## Number of studies combined: k = 18
## 
##                          SMD             95%-CI     z  p-value
## Random effects model 27.8579 [24.5420; 31.1738] 16.47 < 0.0001
## 
## Quantifying heterogeneity:
##  tau^2 = 49.6504 [27.0335; 114.0840]; tau = 7.0463 [5.1994; 10.6810]
##  I^2 = 96.5% [95.5%; 97.3%]; H = 5.37 [4.72; 6.11]
## 
## Test of heterogeneity:
##       Q d.f.  p-value
##  490.22   17 < 0.0001
## 
## Results for subgroups (random effects model):
##                   k     SMD             95%-CI    tau^2     tau     Q   I^2
## Population = YE   2 27.6321 [14.2166; 41.0476]  91.9909  9.5912 54.71 98.2%
## Population = RE   2 29.1938 [21.1505; 37.2370]  31.5726  5.6189 15.92 93.7%
## Population = GI   2 25.5462 [14.0961; 36.9963]  65.9414  8.1204 29.43 96.6%
## Population = MU   2 27.9465 [13.4112; 42.4819] 108.1892 10.4014 60.80 98.4%
## Population = MA   2 33.0233 [27.0389; 39.0077]  16.8248  4.1018 10.21 90.2%
## Population = UM   2 29.5135 [18.2354; 40.7917]  64.3465  8.0216 35.24 97.2%
## Population = KA   2 25.5088 [ 9.9951; 41.0224] 123.5766 11.1165 72.51 98.6%
## Population = VA   2 25.3214 [ 7.2895; 43.3534] 167.5076 12.9425 95.18 98.9%
## Population = AK   2 27.1328 [15.0911; 39.1746]  73.8066  8.5911 44.68 97.8%
## 
## Test for subgroup differences (random effects model):
##                     Q d.f. p-value
## Between groups   2.52    8  0.9607
## 
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau

# run the analysis with EK's code for quick plotting
Pgm_TotalPercwoSchmidt_meta <- metaAnalysisRandomModel(Pgm_TotalPerc_effects %>% filter(Sex == "F" & Lab != "Schmidt"))

p_Pgm_TotalPercwoSchmidt_meta_SE <- Pgm_TotalPercwoSchmidt_meta$summary_effects_random %>%
  dplyr::select(Population, Mstar, SEMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = Mstar + SEMstar, xmin = Mstar - SEMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with SE") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Pgm_TotalPercwoSchmidt_meta$Qtest_het$Q), hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Pgm_TotalPercwoSchmidt_meta$Qtest_het$p), hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
  
p_Pgm_TotalPercwoSchmidt_meta_CI <- Pgm_TotalPercwoSchmidt_meta$summary_effects_random %>% 
  dplyr::select(Population, Mstar, ULMstar, LLMstar) %>% 
  distinct() %>%
  ggplot(aes(x = Mstar, y = 1:length(Population), color = Population)) +
  theme_bw() +
  geom_point(size = 4, shape = 15) +
  geom_errorbarh(aes(xmax = ULMstar, xmin = LLMstar), height = 0) +
  scale_y_continuous(name = "Population", breaks = 1:9, 
                     labels = c("YE","RE","GI","MU","MA","UM","KA","VA","AK")) +
  labs(x = "Summary effect", title = "Pop. summary effect with 95% CI") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank()) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(Q) == %.4g", Pgm_TotalPercwoSchmidt_meta$Qtest_het$Q), hjust=-0.2, vjust=3.5, parse = T) +
  annotate("text", x = -Inf, y = Inf, label = sprintf("italic(p) == %.4g", Pgm_TotalPercwoSchmidt_meta$Qtest_het$p), hjust=-0.2, vjust=5, parse = T) +
  theme(legend.position = "none")
  
p_Pgm_TotalPercwoSchmidt_meta <- ggarrange(p_Pgm_TotalPercwoSchmidt_meta_SE, p_Pgm_TotalPercwoSchmidt_meta_CI)
p_Pgm_TotalPercwoSchmidt_meta

Results III : Trait Correlations tbd

Stress related traits

Temperature related traits

Size related traits

Results III : Qst - Fst tbd

Results IV : Environmental Variables tbd

Results V : Heritability tbd

---

1. @ Schmidt lab, females were allowed to lay eggs for Xh, and viability is calculated as the percentage of individuals that emerged from total number of eggs, per vial.↩︎